Show Summary Details

Page of

PRINTED FROM the OXFORD RESEARCH ENCYCLOPEDIA, ECONOMICS AND FINANCE (oxfordre.com/economics). (c) Oxford University Press USA, 2018. All Rights Reserved. Personal use only; commercial use is strictly prohibited (for details see Privacy Policy and Legal Notice).

date: 13 December 2018

Information, Risk Aversion, and Healthcare Economics

Summary and Keywords

The terms information and risk aversion play central roles in healthcare economics. While risk aversion is among the main reasons for the existence of health insurance, information asymmetries between insured individual and insurance company potentially lead to moral hazard or adverse selection. This has implications for the optimal design of health insurance contracts, but whether there is indeed moral hazard or adverse selection are ultimately empirical questions. Recently, there was even a debate whether the opposite of adverse selection—advantageous selection—prevails. Private information on risk aversion might weigh out information asymmetries regarding risk type and lead to more insurance coverage of healthy individuals (instead of less insurance coverage in adverse selection).

Information and risk preferences are important not only in health insurance but more generally in health economics. For instance, they affect health behavior and, consequently, health outcomes. The degree of risk aversion, the ability to perceive risks, and the availability of information about risks partly explain why some individuals engage in unhealthy behavior while others refrain from smoking, drinking, or the like.

Information has several dimensions. Apart from information on one’s personal health status, risk preferences, or health risks, consumer information on provider quality or health insurance supply is central in the economics of healthcare. Even though healthcare systems are necessarily highly regulated throughout the world, all systems at least allow for some market elements. These typically include the possibility of consumer choice, for instance, regarding health insurance coverage or choice of medical provider. An important question is whether consumer choice elements work in the healthcare sector—that is, whether consumers actually make rational or optimal decisions—and whether more information can improve decision quality.

Keywords: health, information, risk aversion, moral hazard, adverse selection, advantageous selection, health behavior

Introduction

Ever since the birth of healthcare as a subdiscipline of economics, the elements information and risk aversion have played a central role in its economics. This is most obvious in the literature on health insurance markets: risk aversion is—besides potential liquidity constraints (see, e.g., Nyman, 2006)—the major reason for individuals to buy insurance against healthcare costs and income loss due to illness. Yet, it is information asymmetries that are potentially leading to both moral hazard and/or adverse selection which rule out a first-best solution of full-cover insurance and lead to cost-sharing models that imply a welfare loss compared to the first-best world. While theoretically this was treated decades ago in the literature (e.g., Arrow, 1963; Rothschild & Stiglitz, 1976), convincing empirical evidence on the actual degree of information asymmetries and moral hazard came much later—with the famous RAND Health Insurance Experiment (Manning et al., 1987) being the exception—and this is still an ongoing field of research.

Risk aversion and information also directly affect health via health-related behavior. More risk-loving individuals are more likely to have an unhealthy lifestyle. This may also be influenced by different risk perceptions. Yet, information on health risks and the consequences of unhealthy behavior might improve health behavior and, later on, health outcomes.

Information has several dimensions. Apart from the just-discussed information on one’s personal health (relevant in the health insurance market) or on health risks (relevant for health behavior), the role of information is also prevalent in other domains of healthcare economics. One example is patient information on quality of physicians, hospitals, nursing homes, and health plans. While traditionally healthcare systems are highly regulated due to market failures and in order to reach goals that are not achieved by market outcomes alone—such as equity in access to healthcare—most healthcare systems have elements of competition to increase efficiency without compromising equity goals. For competition elements to actually work, consumers need to be in the position to make rational and well-informed choices. Consumer information, thus, plays an important role in making healthcare markets work and enhance welfare. Again, this is theoretically well understood, but whether more or better information is actually used by individuals to improve their decisions on healthcare markets is an empirical question.

This article discusses the empirical economic literature in the field of information, risk aversion, and healthcare economics. It is structured into three parts. The first part discusses implications of information and risk aversion on health insurance markets. The second part sheds light on their impact on individual health behavior. The third part leaves the aspect of risk preferences more implicit and focuses on other facets of information such as consumer information on healthcare providers.

To be more specific, the first section focuses on the health insurance market, probably the most widely studied area in health economics. Still, research in this area is probably more active than ever. On the one hand, this is due to greatly increased data quality and quantity in recent years. On the other hand, the credibility revolution (Angrist & Pischke, 2010) in empirical economics also took place in the area of health economics, and scholars now have an increased awareness of the ingredients necessary to carry out a credible causal analysis—this opens the door for new methods to more convincingly answer “old” questions than has been possible in the past. Research centers around whether there are information asymmetries between health insurance companies and insured individuals at all, whether individuals make use of their private information (e.g., for overconsumption of health services, i.e., moral hazard), or whether private information necessarily leads to adverse selection and not actually to the opposite, advantageous selection—due to risk aversion, healthy individuals could actually have more insurance coverage. All of these ultimately empirical questions are highly relevant for the optimal design of health insurance contracts.

The second section looks at the direct effects of risk aversion and information on health behavior. Both are linked via risk perception, which is potentially formed by risk aversion and information and can be expected to affect habits such as smoking, drinking, eating unhealthy food, and not exercising. Knowledge on the interplay of these characteristics and outcomes should be important for policymakers aiming at improving public health.

The next section restricts the analysis to information and healthcare markets and asks whether more information is actually processed by individuals and used to make better decisions when buying health insurance or deciding for a physician, hospital, or nursing home. Ever-increasing information, also due to the Internet, could either be helpful or harmful if information overload deteriorates decision quality. The final section provides a brief summary.

This article reflects the multitude of impacts information and risk aversion have on healthcare economics. It gives brief snapshots of research in a variety of fields of health economics linked to these terms. It is far from being a complete treatment of these issues and more meant to be an overview of some active strands of the empirical health economic literature in the early 21st century—according to our own subjective assessment.

Information Asymmetries in Health Insurance

Moral Hazard

Moral hazard is defined as “the change in health behaviour and healthcare consumption caused by insurance” (Zweifel & Manning, 2000). This may imply excess demand for health services. In a world with risk-averse individuals and without moral hazard, full-cover insurance, that is, insurance that covers all treatment costs without any patient cost-sharing, is optimal. Yet this first-best solution is not feasible when moral hazard is present. In such a case, the introduction of cost-sharing can be a welfare-enhancing second-best solution (see Zweifel, Breyer, & Kifmann, 2009). The optimal amount of cost-sharing is unclear from a theoretical perspective and depends on the degree of moral hazard (or the price-elasticity of healthcare demand). Determining this is an empirical question.1

Due to a fundamental selection problem it is challenging to identify the price elasticity, that is, the causal effect of insurance coverage on healthcare utilization: healthy individuals or those who expect less healthcare utilization tend to buy insurance contracts with less coverage. The mere empirical observation of lower healthcare costs induced by individuals with less insurance coverage, thus, does not immediately imply the existence of moral hazard and the need for cost-sharing.

RAND Health Insurance Experiment and Oregon Health Insurance Experiment

The probably most convincing way to determine the causal effects health insurance coverage has on the demand for medical services is a randomized experiment in order to rule out self-selection problems. The largest and best-known randomized experiment in the health economic literature is the RAND Health Insurance Experiment (Manning et al., 1987). Between November 1974 and February 1977, 5,809 individuals in six U.S. cities were randomly assigned to different health insurance plans. The plans varied in the rate of coinsurance (from 0 to 95% of out-of-pocket expenditures) and upper limits of annual out-of-pocket payments (between 5 and 15% of annual family income, capped by $1,000). The results concerning demand for medical services (number of treatments and expenditures caused by the treatments) show an incentive effect. Individuals without any co-insurance caused the highest expenditures due to the highest demand for healthcare. The demand decreased with higher amounts of copayments. Yet it is somewhat inelastic, with an average elasticity of around −0.2 (Keeler & Rolph, 1988). Outpatient visits were found to be more elastic than inpatient visits. The findings for health outcomes are less pronounced, with only a handful of utilized health measures showing a statistically significant change due to insurance coverage. This might have also been partly due to small sample sizes, as 5,800 individuals were randomized into 14 different health plans, and potentially also due to sample attrition (Aron-Dine, Einav, & Finkelstein, 2013). Moreover, only short-term effects were identified, and it might well take some time for health effects to materialize.

Another—and much more recent—piece of credible evidence on the impact of insurance in the United States comes from the Oregon Health Insurance Experiment (Finkelstein et al., 2012). Here, an institutional coincidence was exploited that randomized individuals into health insurance coverage. The state of Oregon had at one time a Medicaid program to insure low-income adults who were not eligible for Medicaid through other means. However, this program was closed in 2004 due to budget shortfalls. In 2008, Oregon had financial resources to cover an additional 10,000 uninsured in the program. As the number of applicants greatly exceeded the available spots, a lottery assigned the spots to the applicants (not all successful applicants eventually signed up). This procedure was used by Finkelstein et al. (2012), who analyze a large set of administrative as well as survey data on the applicants. They find that Medicaid coverage increased the number of doctor visits by about 35% in the first year. It also strongly increased emergency room use (Taubman, Allen, Wright, Baicker, & Finkelstein, 2014). Drug spending increased about 15% and overall healthcare spending by 25%. As in the RAND Health Insurance Experiment, short-run health effects were less pronounced. It was mainly subjective, and mental health measures increased while objective and physical health measures changed less (or not at all).

These highly internally valid estimates of the effect of health insurance on healthcare utilization and health might have come at the cost of external validity. Strictly speaking, the findings only hold for the group of low-income uninsured residents of Oregon interested in obtaining public coverage. Moreover, long-run effects might be different than the identified short-term effects as these could be driven by pent-up demand of uninsured individuals. Nevertheless, this study probably delivers the most credible and convincing findings of how (certain) individuals react to health insurance coverage that is currently available.

Natural Experiments, Discontinuities, and Instrumental Variables

While randomized experiments are often seen as the gold standard in the social sciences, there are other ways to elicit credible causal effects. These are through natural experiments (mainly by policy changes) or institutional peculiarities that generate discontinuity that can be exploited by researchers. As one example, Shigeoka (2014) uses an institutional detail in Japan: when people turn 70, their cost-sharing rate strongly drops. In a regression discontinuity design, he finds that outpatient and inpatient care increases at the age threshold while there is, again, little impact on health outcomes.

By use of a difference-in-differences estimator, Ziebarth (2010) exploits an increase in copayments for convalescent care programmes for publicly (but not privately) insured in Germany to find a price elasticity for medical rehabilitation programmes aimed at preventing work disability of about −0.3. Other European evidence comes from Chiappori, Durand, and Geoffard (1998), Cockx and Brasseur (2003), as well as Boes and Gerfin (2016), who find price elasticities comparable to the one in the RAND experiment. In contrast, Schreyögg and Grabka (2010) do not find any effects of a copayment increase in Germany on the number of physician visits.

More evidence from the United States, focusing on elderly individuals, stems from Card, Dobkin, and Maestas (2008, 2009) and Chandra, Gruber, and McKnight (2010). The former find that an exogenous increase in insurance coverage at the age of 65 (when some previously uninsured qualify for Medicare) increases healthcare utilization and also has modest positive health effects. The latter detect very small effects of an increase in copayments for physician office visits and prescription drugs in a supplemental Medicare insurance policy for Californian civil servants. However, they find an “offset” effect as hospital utilization is increased.

Coming closer to this article’s topic of “information, risk aversion, and healthcare economics,” Schmitz (2012) does not use a natural experiment but the individual degree of risk aversion to instrument the health insurance choice in Germany. In a latent-class model, he finds that deductibles have no effect on healthcare utilization for the high users of healthcare while they reduce the number of doctor visits for the low users.

Dynamic Effects and Nonlinear Price Schedules

In revisiting the RAND Health Insurance Experiment, Aron-Dine et al. (2013) take issue with the claim that the price elasticity of healthcare demand is about −0.2. Given that, typically, health insurance contracts are highly nonlinear, individuals face different prices throughout the year, implying that there are different price elasticities—or, put differently, they face elasticities with respect to different prices. As a simple example, consider an individual with an insurance contract with deductible, copayment, and stop-loss elements. The first 500 euros one incurs per year need to be paid out of pocket. After this threshold health insurance steps in, but with a copayment of 10%. Annual costs exceeding 2,000 euros are fully borne by the insurance company. Thus, depending on the time of the year and how much medical care an individual has already consumed, the “spot” out-of-pocket price for 1 euro of health expenditure can either be 1, 0.1, or 0. The “end-of-year” price can also be 1, 0.1, or 0, and the average price throughout the year is something in between.

Recent literature explicitly focuses on these nonlinearities in insurance contracts and, first of all, asks whether individuals take the dynamic incentives into account when deciding how much medical care to consume—that is, whether they are myopic and only react on the (current) spot price or whether they realize that the (future) end-of-year price is also affected by today’s behavior. Aron-Dine, Einav, Finkelstein, and Cullen (2015) indeed find that, in Medicare Part D in the United States, individuals not only respond to the spot price but also take the future price into account. Other studies from the United States that explicitly pay attention to nonlinear insurance contracts are Vera-Hernandez (2003), Bajari, Hong, Park, and Town (2011), Kowalski (2015), Dalton (2014), and Dalton, Gowrisankaran, and Town (2015).

In a study using European data, Gerfin, Kaiser, and Schmid (2015) use detailed expenditure data from a Swiss health insurer and exploit the fact that deductibles are reset at the beginning of each year—thus, also taking into account different prices for the same individual throughout the year. They find that healthy individuals (with higher deductibles) respond much more strongly to financial incentives than do less healthy ones with smaller deductibles. Yet here, and in other recent literature on this issue, it is not fully clear whether increased health expenditure, once the deductible is hit, is actually evidence for moral hazard or rather for intertemporal substitution (or both). Individuals who expect their deductible to be reset the next year might consume medical services in the current year that they would have consumed next year anyway. The work of Cabral (2017) and Einav, Finkelstein, and Schrimpf (2015) underline the importance of not isolating the analysis to single years but looking at horizons longer than one year.

Adverse and Advantageous Selection2

The other problem of information asymmetry in health insurance is adverse selection. Standard insurance models with asymmetric information like the Rothschild-Stiglitz model assume that individuals who are bad risks choose higher coverage than do good-risk individuals (Rothschild & Stiglitz, 1976). Private individual information on the true risk type prevents insurance companies from perfectly calculating proper insurance premiums for all risk types that might drive the good risks out of the market, that is, there is adverse selection in the insurance market.3 The model predicts a positive correlation between insurance coverage and the occurrence of the insured risk conditional on the information of the insurance provider.

An empirical test of the positive correlation between insurance coverage and the occurrence of the risk (the “positive correlation test”) conditional on all information observed by the insurance company can be seen as one possible indication of the presence of asymmetric information in an insurance market (Chiappori & Salanie, 2001). Rejecting the hypothesis of no asymmetric information in favor of a positive correlation is evidence for either moral hazard, adverse selection, or both together, which are usually difficult (though not impossible) to separate.

However, many empirical applications find no evidence for adverse selection in different insurance markets like markets for long-term care, Medigap insurance, or life insurance (see Cutler, Finkelstein, & McGarry, 2008, for an overview). In some markets there can even be found a negative correlation between insurance coverage and experience of risks. One explanation for this finding is that not only do individuals have private information on their risk type (possibly leading to adverse selection) but that preferences like risk aversion also shape their demand for insurance and the probability of experiencing the insured risk. A more risk-averse individual might demand more insurance and at the same time try to minimize the probability of occurrence of the risk. Risk aversion is also unobserved by the insurer, but in this case this information asymmetry does not lead to adverse selection, it is rather a source of the opposite, namely advantageous selection. Because private information is not one-dimensional (only on risk type) but multi-dimensional (also on risk preferences), potential sources of adverse and advantageous selection may net out and the overall effect is not clear a priori. While the Rothschild-Stiglitz model predicts a positive correlation of risk type and insurance coverage, de Meza and Webb (2001), who explicitly allow for differences in risk preferences among individuals, also allow for equilibria that exhibit a negative correlation between risk and insurance coverage.

Therefore, the positive correlation test as a test for asymmetric information is invalid (Finkelstein & Poterba, 2014). Failure to reject the hypothesis of no asymmetric information can arise if there are sources of adverse and advantageous selection that partly cancel each other out. Finkelstein and Poterba (2014), thus, propose another test on information asymmetry, the “unused observables test.” The existence of only one variable that is not used by the insurer to calculate the risk classification of the insured, but which is correlated with both the insurance choice and the risk of the insured loss (conditional on the observed variables), is evidence of asymmetric information. Finkelstein and Poterba (2014) find for the United Kingdom that the place of residence is correlated with both purchasing annuities and the annuitant’s mortality but not being used for insurer’s risk calculation, thus leading to adverse selection.

Cutler et al. (2008) review the empirical literature and find that whether there is more or less insurance of high-risk individuals depends on the type of insurance market. While in acute care and annuity markets high-risk individuals usually buy more insurance (as the standard model predicts; see Cutler & Zeckhauser, 2000, or Finkelstein & Poterba, 2004), the opposite is true in the case of life insurance, long-term care, and Medigap markets (Cawley & Philipson, 1999; Fang, Keane, & Silverman, 2008; Finkelstein & McGarry, 2006). Doiron, Jones, and Savage (2008) also find evidence for advantageous selection for private supplementary health insurance in Australia, which, however, is in contrast to the previous findings in the American market for acute health insurance.

Cutler et al. (2008) furthermore find that risky behavior like smoking, alcohol abuse, not doing preventive care, not using a seat belt, or holding a risky job is negatively correlated with holding insurance in all the mentioned markets. However, these risk-tolerant individuals have higher claims for long-term care insurance and life insurance and lower claims for annuities. In contrast, the results for Medigap and acute health insurance are mixed. Fang et al. (2008) show that not only attitudes toward risks as measured by these proxies may be important but that there are also other possible variables that correlate with insurance and experience of risk. They find that cognitive ability and wealth are sources of advantageous selection in the Medigap market.

Buchmueller, Fiebig, Jones, and Savage (2013) do not find a positive correlation between risk and insurance coverage in Australia. They suggest—but can only provide indirect evidence—that this is due to risk aversion as a source of advantageous selection. Schmitz (2011) uses German data and a self-stated measure for risk aversion to directly test whether risk preferences can be a source of advantageous selection. He finds risk-averse men to be more likely to buy private supplementary insurance for hospital visits but to need the extra coverage less due to a smaller number of hospital visits. No such relationship is found for women, however.

Risk Aversion, Risk Perception, and Information as Determinants of Health Behavior

Risk Aversion and Health Behavior

Individual health is strongly influenced by human behavior. But why do some people behave unhealthily while others refrain from smoking or drinking and instead exercise, eat healthy food, and use preventive care? Apart from variables such as educational background and socio-economic status, preferences and personality traits seem to be important explanatory factors. One such preference is the attitude toward risks.

In empirical studies on the effects of risk aversion on health behavior, a major challenge is the actual elicitation of individual risk aversion. Approaches can broadly be categorized into four blocks: (1) subjective assessments of individuals’ own risk preferences, (2) hypothetical gambles, (3) incentivized laboratory or field experiments, and (4) preference parameters estimated from structural models using data on observed behavior. There is no single best method to elicit risk preferences; all approaches have advantages and disadvantages. In the following we will separately discuss studies in the field of risk aversion and health behavior that used these different methods and, thereby, mention their strengths and weaknesses. Another way to differentiate studies is by domain of risk aversion. Some studies employ a general notion of risk aversion while others focus on risk aversion with respect to health issues.

Subjective Assessments

Dohmen et al. (2011) investigate survey responses about the self-reported willingness to take risks in the German Socio-Economic Panel and their potential to predict risky behavior. Respondents were asked to respond to the question “Are you generally a person who is willing to take risks or do you try to avoid taking risks?” on an 11-point scale between 0 (not at all willing to take risks) and 10 (very willing to take risks). Another question directly referred to risk willingness with respect to health. Dohmen et al. (2011) find risk preferences to be strongly but imperfectly correlated across domains. Hence, a question about general risk taking serves as the best all-round predictor of behavior, but context-specific questions might better explain the power of actions within a particular domain. For instance, smoking is best predicted by the question about the willingness to take risks in health matters. If the latter increases by one standard deviation, the probability of smoking rises by 20%.

Using a similar variable on general risk aversion, van der Pol, Hennessy, and Manns (2017) find that, in contrast to time preferences, risk aversion does not seem to play a role in adherence with doctoral advice to change health behavior, such as dietary changes and changes in physical activity.

A major advantage of this measure of risk aversion is its simplicity and, thus, it is easy to incorporate into large-scale surveys. Yet different reporting styles and, in general, its subjective nature renders this measure unusable for some researchers. Biases might result due to self-serving behavior, strategic motives, or a lack of concentration when answering questions (Dohmen et al., 2011).

Hypothetical Lotteries

Barsky, Juster, Kimball, and Shapiro (1997) use hypothetical gambles over lifetime income to elicit risk preferences. They find a statistically and quantitatively significant relationship between risk aversion and smoking as well as drinking. However, using a similar approach, Sato and Ohkusa (2003) do not find a significant effect of risk attitude on smoking initiation. Van der Pol and Ruggeri (2008) more strongly support the domain-specific risk aversion. Based on the results of gambles involving the risk of immediate death, differing remaining life-years, and severity of illness, they emphasize that individuals seem to be risk-averse with regard to the risk of immediate death but risk-seeking regarding the other health risks. This variability in the attitude toward risk stands in contrast to the traditional economic assumption of a uniform risk preference and complicates the examination of the impact of risk aversion on health behavior. However, in a follow-up study that improved elicitation of risk preferences, Ruggeri and van der Pol (2012) find individuals to be risk-averse in the gambles on life-years as well as quality of life. Hypothetical lotteries are a much more sophisticated and detailed way to elicit risk preferences than an assessment on a one-dimensional scale, yet they still measure stated preferences only, which might differ from true (revealed) preferences.

Incentivised Lotteries

The predominant view among experimental economists seems to be that risk preferences should be measured by observing actual behavior instead of relying on stated preferences (e.g., Charness, Gneezy, & Imas, 2013; Holt & Laury, 2002). This is often confirmed by incentivized experiments in the laboratory.

For example, Anderson and Mellor (2008) conduct the choice experiment designed by Holt and Laury (2002). In addition, health-related risky behaviors are reported through a survey. The results reveal a negative relationship of risk aversion and the probability of smoking, overweight, and seat belt non-use as well as the likelihood of heavy drinking. Yet the results of other studies using incentivized experimental designs are mixed. Blondel, Loheac, and Rinaudo (2007) find a negative relationship between risk aversion and risky behavior in the case of drug consumption, whereas the estimates of Harrison, Lau, and Rutstrom (2010) do not show a significant negative association or even a positive one in the case of smoking. The downside of this approach concerns the validity of results outside the laboratory. Moreover, experimental studies are hard to conduct with a large, representative sample.

In an attempt to compare results from surveys and incentivized experiments, Dohmen et al. (2011) combine both methods and find a large degree of conformity regarding the derived risk attitude. Thus, survey data seem to be able to reveal risk preferences that affect behavior.

Preference Parameters From Structural Models

The most traditional method to estimate preference parameters such as risk attitudes is by exploiting information from observed choices of economic agents using structural models. Thus, there is a large literature (see the survey by Barseghyan, Molinari, O’Donoghue, & Teitelbaum, 2018) on this. Yet risk preferences are usually not related to health behavior in that literature and are, thus, outside the scope of this article. The major advantages of this approach are the arguably higher external validity compared to laboratory experiments and higher internal validity compared to stated preferences. The major drawback seems to be that structural models are often highly parameterized and depend on a number of assumptions. Thus, it seems most promising to combine structural models with natural experiments in order to credibly estimate parameter such as brought forward by Chetty (2009) in his “sufficient statistics” approach.

While Dohmen et al. (2011) show the comparability of survey measures and those derived from incentivised experiments, Fossen and Glocker (2017) go one step further and compare survey measures of risk aversion with those estimated from structural models and observed actual behavior. While not in the domain of health economics but education economics, they find consistency between stated and revealed preferences, again backing the validity of the survey instrument.

Impact of Health on Risk Preferences

Traditionally, risk preferences are expected to be constant over an individual’s lifetime (Stigler & Becker, 1977). Nevertheless, it seems plausible that, despite its comparability to rather fixed personality traits, risk attitudes can alter due to major life events, like the onset of severe health conditions. Thus, risk aversion does not only affect health (behavior), but health status may affect risk preferences. This interdependency complicates the identification of unidirectional causal effects. Research studies concerning the analysis of changes in risk tolerance mostly observe health effects. Sahm (2012) finds no clear evidence, while Gloede, Menkhoff, and Waibel (2015) and Schurer (2015) do. However, it is often impossible to infer causality from the observed heterogeneities in risk preferences. In order to take selection and reverse causality into account, Decker and Schmitz (2016) apply regression-adjusted matching and control for possibly confounding characteristics such as pre–health shock risk aversion. They find effects of health shocks (as measured by a strong decline in hand-grip strength over time) on the willingness to take risks. More precisely, individual risk aversion increases by about 9 to 11% of a standard deviation in the risk preference. Hence, the traditional (implicit) assumption of constant risk preferences seems not to hold.

Risk Perception and Health Behavior

In addition to the willingness to take health risks, the perception of risks is a potential further determinant of health behavior. The stronger the perceived risks, the less likely a person is to engage in risky behaviors. Cutler and Glaeser (2005) show that individuals who believe that smoking or drinking is very harmful have a lower probability to smoke or drink. Lundborg and Andersson (2008) examine the behavioral impact of the perceived risk of mortality and addiction of smoking through the use of an alcohol and drug survey conducted in Swedish schools. They find that stronger risk perceptions in both categories are associated with a decrease in the probability of smoking. Contrary to the general notion that females dislike risks more than males, no gender difference in the magnitude of the reaction to a changed risk perception is determined.

Gerking and Khaddaria (2012) consider the aspect that the effect of risk perception on health behavior is not homogeneous. They analyze the impact of perceived risk on smoking, conditional on the perception of the difficulty to quit smoking and the beliefs of the immediacy of harm. The authors show that for people who believe that it is difficult to stop smoking and that health effects occur relatively quickly after smoking initiation, an increase in perceived risk has a deterrent effect on smoking. In contrast, smoking status is not affected by risk perceptions if the individual holds the opposite beliefs. Thus, some people engage in harmful behavior because they overestimate their ability to cease before bad consequences set in.

Besides the false assessment of the risk of addiction, it is possible that smokers take risks because of an underestimation of the probabilities of harm (Sloan & Platt, 2011). Hammar and Johansson-Stenman (2004) sent a questionnaire to 935 Swedish smokers asking for the price they would be willing to pay for a totally risk-free cigarette. The respondents put relatively low values on a lost life-year, indicating an underestimation of the health risks of smoking. Furthermore, they seem to be too optimistic regarding their ability to stop smoking, equaling an underestimation of the risk of addiction. Nevertheless, the empirical evidence of the correctness of risk assessment of health risk takers is heterogeneous. Sloan and Platt (2011) determine the correctness of risk perceptions of adolescents by comparing subjective and objective probabilities. Unfortunately, the survey does not entail information about specific beliefs in the health domain. Instead, probabilities of dying (irrespective of the cause) and of being a crime victim are analyzed. The results point to a general pessimistic perception of risks and a more than objectively detectable increase in the subjective probability of mortality risk after smoking initiation. This is suggestive of an overestimation of smoking-related health risks. Accordingly, the effectiveness of health information campaigns that only publicize probabilities of harm can be questioned. Khwaja, Silverman, Sloan, and Wang (2009) try to quantify the accuracy of smoker’s risk perceptions with respect to smoking-related morbidities, physical difficulties, and prospective death. Regression estimates of the difference in subjective and objective probabilities indicate that individuals have on average quite correct perceptions of the risk of surviving to age 75 and the occurrence of physical difficulties, and this holds irrespective of the smoking status. Furthermore, non-smokers and smokers tend to be highly pessimistic with respect to the onset of chronic diseases. However, the studied age group is restricted to people between 50 and 70 years. Khwaja et al. (2009) mention that younger individuals who start smoking are more likely to underestimate health risks. Additionally, smokers of all ages appear to underestimate the risk of addiction.

Nevertheless, the estimation of the impact of risk perception on health behavior requires caution due to at least three different reasons (Lin & Sloan, 2015): disregarded factors that affect both risk perceptions and health behavior, for example, preferences for a good health state, cause an omitted variable bias; beliefs about smoking harms may be measured with error; and reverse causality can be at play. The latter possibly occurs because of ex post rationalization to justify unhealthy behavior. To avoid biased estimates, Lin and Sloan (2015) conduct two-stage least-squares regressions. The authors use instrumental variables that capture the proximity to a neighbor who experienced a smoking-related health shock. The knowledge of a neighbor’s lung cancer influences a smoker’s risk perception and, subsequently, a smoker’s behavior. A one-unit rise in risk perception measured on a five-point scale increases the probability of smoking cessation within two years after the cancer diagnosis by 4.3 percentage points.

If people behave unhealthily because they are not able to correctly assess the risk of their actions, education may help to enhance health and to reduce related costs by improving cognitive abilities. Indeed, numerous studies confirm the existence of a positive effect of schooling on health (e.g., Brunello, Fort, Schneeweis, & Winter-Ebmer, 2016; Cutler & Lleras-Muney, 2010; Jensen & Lleras-Muney, 2012; Kemptner, Jürges, & Reinhold, 2011; Kenkel, 1991) as well as between college education and health behavior (Currie & Moretti, 2003; de Walque, 2007; Grimard & Parent, 2007; Kamhöfer, Schmitz, & Westphal, 2017). However, the mechanisms are not yet understood, and it is not clear whether an improved processing of relevant information (and, thus, more accurate risk perception) is related to the positive effect of education on health.

Information and Health Behavior

It is possible that health risk-takers engage in risky behavior not only because of a risk-loving attitude or a misperception of potential harm but because of a lack of information about the consequences and their probabilities of occurrence. Individual risk perceptions can be seen as the outcome of the calculation of a weighted average of different information sources (Lundborg & Andersson, 2008). Hence, the relations between perceptions of risk and information about risks are interdependent.

While the general effect of education on health behavior has briefly been mentioned, we now turn to the effects of more specific interventions to increase health knowledge and literacy. Some studies use survey questions to determine the degree of health information or health (risk) knowledge (e.g., Kan & Tsai, 2004; Sato & Ohkusa, 2003). However, it can be questioned if knowledge can be treated as exogenous, especially when measured by a survey (Kenkel, 1991). Analogous to the estimation of the effect of risk perception on health behavior, omitted variable bias and reverse causality may be prevailing. The most credible studies in this field exploit field experiments in order to exogenously vary information to achieve credible causal effects.

As one example, Wisdom, Downs, and Loewenstein (2010) conducted a field experiment at a fast-food sandwich chain to investigate the effect of information on food choice. The provision of calorie information results in a decreased calorie intake. Daily calorie recommendations also appear to be effective, but only for non-overweight individuals. This could be the case because dieters may impose strict calorie restrictions upon themselves in order to eat less (Downs, Loewenstein, & Wisdom, 2009). Calorie recommendations that are too high may thus be ineffective or even increase calorie intake.

Regarding another kind of health behavior, unprotected sex, Dupas (2011) conducts a randomized field experiment in Kenya to evaluate the effect of HIV risk information. The random assignment of schools to the treatment group, in which teenagers receive information about the relative risk of a HIV infection by partner’s age, should ensure that treatment and control group differ only with respect to the treatment. The provision of information decreased teen pregnancy by 28%. The latter serves as proxy for unprotected sex.

Apart from experimental settings, there are other sources of (arguably) exogenous variation in information. As an example, a health shock can be interpreted as health information. Based on the individual-level data, Bünnings (2016) uses health events like physical health problems, mental disorders, and accidents as proxies for new health information and examines their effect on the decision to quit smoking. He finds that health problems, in particular those of the physical type, motivate individuals to stop smoking.

To conclude, various empirical studies suggest that health behavior is indeed influenced by information.

Information and Its Impact on Consumers and Providers of Medical Services

Even though healthcare systems are necessarily highly regulated throughout the world, all systems allow for some market elements. These typically include the possibility of consumer choice, for instance, regarding health insurance coverage or the choice of medical providers. An important question is whether consumer choice elements work in the healthcare sector—that is, whether consumers actually make rational or optimal decisions—and whether more information can improve decision quality or not. Suboptimal choices despite the provision of full information stands in contrast to the standard economic theory: one would expect that the matching quality between consumer preferences and health plans or medical providers increases. In addition, welfare should be enhanced because of greater price and quality competition (McWilliams, Afendulis, McGuire, & Landon, 2011). Yet empirical evidence shows that this is not always the case.

Information and Consumer Choice in Health Insurance

In many countries, individuals can choose among different health plans to protect themselves against the costs of health expenditures. The Internet serves as information pool and as such should facilitate efficient choices in accordance with individual preferences. In the United States, the Patient Protection and Affordable Care Act of 2010 led to the mandatory creation of health insurance exchanges. Via an online marketplace, different health plans can be compared. Yet individuals do not seem to transform this into optimal choices for two main reasons: insurance contracts are often too complex and differences in health plan characteristics are sometimes not salient enough.

Complexity

Insurance choices of seniors under the Medicare Part D Prescription Drug plan are well studied. According to Abaluck and Gruber (2011a) and Heiss, Leive, McFadden, and Winter (2013), many individuals do not choose optimally because, inter alia, they overemphasize premiums compared to plan generosity in their decision process. Additionally, plans are chosen that provide worse risk protection at higher costs than alternative plans, and the inefficiencies hold irrespective of age, gender, predictability of drug demand, or size of predicted drug expenditures (Abaluck & Gruber, 2011b). However, sicker individuals tend to have more difficulties identifying the optimal health plan. This also holds for individuals with cognitive deficits (McWilliams et al., 2011).

Inefficiencies can be attributed to the complexity of a health plan choice (Sinaiko & Hirth, 2011). Frank and Lamiraud (2009) find that an increase in the number of available health plans in Switzerland actually reduces the likelihood of switching. Laboratory experiments can be used to mimic the selection of health plans and the complex decision environment. Results show that a rise in the number of alternatives leads to an increase in decision-making time, reduces the used fraction of the available information set, and deteriorates decision quality (Schram & Sonnemans, 2011). An objective ranking of available insurance plans enables the determination of optimal decisions (Besedeš, Deck, Sarangi, & Shor, 2012). It is striking that as the number of available options increases, the probability to choose the optimal plan shrinks.

Other empirical studies provide valuable insights regarding the consumer choice of health insurance plans. Ericson and Starc (2012) analyze a complete set of insurance choices made through the Massachusetts’s health insurance exchange. The resulting information or choice overload causes that many individuals do not behave rationally and instead rely on the heuristic to choose the cheapest but also least generous health plan. For the German health insurance market, Bünnings, Schmitz, Tauchmann, and Ziebarth (2018) show that many individuals choose plans that are worse according to price and non-price attributes at the same time compared to other plans in the market. This holds although heavy federal regulation ensures a simple choice architecture for insured individuals. Enrollees in dominated plans are, on average, older, less educated, and less healthy. Bhargava, Loewenstein, and Sydnor (2017) challenge the view that complexity is the major driver of bad quality decisions. They find that “employees’ lack of understanding basic health insurance concepts” (p. 2017) explain why the majority of individuals in their sample of U.S. employees choose dominated health plans.

Salience

Schmitz and Ziebarth (2017) find that a simple change in price setting can help individuals to switch to cheaper health plans with the same benefit package—that is, help individuals not to leave money on the table. In 2009, price differences were made more salient, making it easier to process information about differences. Before the reform, differences were expressed in percentage points of contribution rates (i.e., percent payroll tax differences, as this is how health insurance premiums are paid in the German social health insurance). Afterwards, price differences between German health plans where expressed in absolute euro values. This strongly increased the individual’s willingness to switch from more expensive to cheaper plans.

Information and Consumer Choice of Hospitals and Nursing Homes

The dissemination of information mitigates informational asymmetries in the healthcare market. For this purpose, public report cards provide information about the quality and costs of physicians and medical institutions. The history of healthcare quality measurement and reporting reaches back to the 19th century when Florence Nightingale published mortality rates of hospitals during the Crimean War (Nightingale, 1858). However, report cards as we know them today are rooted in reporting activities of public agencies such as in New York and Pennsylvania, which released cardiac surgical outcomes in the beginning of the 1990s (Sinaiko, Eastman, & Rosenthal, 2012). Since then, besides the United States, other Western countries have started to publish quality and cost information for the healthcare system. While in the last century, interested consumers were only able to find information at the respective institution or through press releases, the Internet achieves a high degree of quality transparency nowadays. Most quality-measurement systems make report cards accessible online (Bates & Gawande, 2000).

Several studies show that better-reported quality goes along with a higher number of patients for both physicians and hospitals (Cutler, Huckman, & Landrum, 2004; Pope, 2009; Varkevisser, van der Geest, & Schut, 2012). Wang, Hockenberry, Chou, and Yang (2011) emphasize that performance disclosure implies a reduction in the demand for services of poorly performing or unrated surgeons, while that of high-quality surgeons remains unaffected. Epstein (2010) asks if physicians changed their referrals to cardiac surgeons according to mortality information given in published report cards but does not find any effects. Referring physicians seem to correctly assess the relative performance of surgeons even without report cards.

Decreased informational asymmetries increase competition in the healthcare market. Consequently, providers of healthcare services have an incentive to improve their quality. Cutler et al. (2004) and Laschober, Maxfield, Felt Lisk, and Miranda (2007) indeed find substantial quality improvements in hospitals after performance measures are publicized. However, Dranove, Kessler, McClellan, and Satterthwaite (2003) emphasize selection behavior of hospitals because report cards create incentives for surgeons to avoid the treatment of patients with complex diseases. As a consequence, sicker individuals are likely to be worse off.

Nursing Homes

Report cards are also released in the nursing home market. Schmitz and Stroka (2014) as well as Werner, Norton, Konetzka, and Polsky (2012) analyze the impact on choice of nursing homes but find no or only minor effects of reported quality on demand for homes. Decisions are mainly based on prices and distances (Schmitz & Stroka, 2014). Nevertheless, studies show that nursing homes react to the increased transparency with quality enhancements, albeit the improvements are rather small (Park & Werner, 2011) and exist mainly for low-quality nursing homes (Clement, Bazzoli, & Zhao, 2012; Herr, Nguyen, & Schmitz, 2016) and especially in highly competitive markets (Grabowski & Town, 2011). According to Werner, Konetzka, and Kruse (2009), even unreported quality of care improved due to report cards. Park, Konetzka, and Werner (2011) analyze the performance and financial outcome of a great number of Medicare-certified nursing homes under public reporting. The results of a facility-level fixed-effects model reveal that reported quality improvements go along with increased profit margins and revenues. Thus, all in all, the provision of information seems to be beneficial for consumers as well as for providers of healthcare services.

In addition to the provision of quality information, information gaps are diminishing in the United States because governments, insurers, and companies provide online price information. Lieber (2017) investigates data from a large U.S. firm that grants a subset of their employees access to information about prices charged by specific medical providers. Difference-in-differences estimates indicate that access to price information decreases payments by 1.6% on average without remarkable reductions in quality.

Impact of Digitalization on HealthCare Economics

In times of digitalization and “big data,” a vast quantity of information is available, which can not only be used by consumers to decide which health plan, hospital, or nursing home to choose but also affects the work of medical providers. An increasing number of studies deals with this issue (see, e.g., Wagner, Hu, & Hibbard, 2001; Miller & Tucker, 2014; Agha, 2014; Javitt, Rebitzer, & Reisman, 2008). This topic will most likely be at the center of empirical health economics in the upcoming decades, and thus we refrain from providing a complete synthesis of the literature that at the beginning of the 21st century is in its infancy.

Conclusion

The terms information and risk aversion play central roles in healthcare economics. Whereas risk aversion is among the main reasons for the existence of health insurance, information asymmetries between insured individual and insurance company potentially lead to moral hazard or adverse selection. This has implications for the optimal design of health insurance contracts, but whether there is indeed moral hazard or adverse selection are ultimately empirical questions. Recently, there was even a debate whether the opposite of adverse selection—advantageous selection—prevails. Private information on risk aversion might offset information asymmetries regarding risk type and lead to more insurance coverage of healthy individuals (instead of less insurance coverage in adverse selection).

Information and risk preferences are important not only in health insurance but more generally in health economics. For instance, they affect health behavior and, consequently, health outcomes. The degree of risk aversion, the ability to perceive risks, and the availability of information about risks partly explain why some individuals engage in unhealthy behavior while others refrain from smoking, drinking, or the like.

Finally, the widespread use of the Internet enables the fast dissemination of information. The question arises whether more information results in a better decision quality when choosing a health plan or deciding on a physician, hospital, or nursing home. Additionally, medical providers are affected by ongoing digitalization.

Further Reading

Angrist, J. D., & Pischke, J. (2009). Mostly harmless econometrics: An empiricist’s companion. Princeton, NJ: Princeton University Press.Find this resource:

Barseghyan, L., Molinari, F., O’Donoghue, T., & Teitelbaum, J. C. (2018). Estimating risk preferences in the field. Journal of Economic Literature.Find this resource:

Cawley, J., & Ruhm, C. J. (2016). The economics of risky health behaviors. In M. Pauly, T. McGuire, & P. Pita Barros (Eds.), Handbook of health economics (Vol. 2, pp. 95–200). Amsterdam, The Netherlands: Elsevier.Find this resource:

Chetty R., & Finkelstein, A. (2013). Social insurance: Connecting theory to data. Handbook of Public Economics, 5(3), 111–193.Find this resource:

Einav, L., & Finkelstein, A. (2011). Selection in insurance markets: Theory and empirics in pictures. Journal of Economic Perspectives, 25(1), 115–138.Find this resource:

Gaynor, M., & Town, R. J. (2016). Competition in health care markets. In M. Pauly, T. McGuire, & P. Pita Barros (Eds.), Handbook of health economics (Vol. 2, pp. 499–638). Amsterdam, The Netherlands: Elsevier.Find this resource:

Lieber, E. M. J. (2017). Does it pay to know prices in health care? American Economic Journal: Economic Policy, 9(1), 154–179.Find this resource:

Ziebarth, N. R. (2017). Social insurance and health. In B. H. Baltagi & F. Moscone (Eds.), Health Econometrics (Contributions to Economic Analysis) (pp. 57–84). Bingley: Emerald Publishing.Find this resource:

References

Abaluck, J., & Gruber, J. (2011a). Choice inconsistencies among the elderly: Evidence from plan choice in the Medicare Part D program. American Economic Review, 101(4), 1180–1210.Find this resource:

Abaluck, J., & Gruber, J. (2011b). Heterogeneity in choice inconsistencies among the elderly: Evidence from prescription drug plan choice. American Economic Review, 101(3), 377–381.Find this resource:

Angrist, J., & Pischke, J.-S. (2010). The credibility revolution in empirical economics: How better research design is taking the con out of econometrics. Journal of Economic Perspectives, 24, 3–30.Find this resource:

Aron-Dine, A., Einav, L., & Finkelstein, A. (2013). The RAND health insurance experiment, three decades later. Journal of Economic Perspectives, 27(1), 197–222.Find this resource:

Aron-Dine, A., Einav, L., Finkelstein, A., & Cullen, M. (2015). Moral hazard in health insurance: Do dynamic incentives matter? Review of Economics and Statistics, 97(4), 725–741.Find this resource:

Agha, L. (2014). The effects of health information technology on the costs and quality of medical care. Journal of Health Economics, 34, 19–30.Find this resource:

Anderson, L. R., & Mellor, J. M. (2008). Predicting health behaviors with an experimental measure of risk preference. Journal of Health Economics, 27(5), 1260–1274.Find this resource:

Arrow, K. (1963). Uncertainty and the welfare economics of medical care. American Economic Review, 53, 941–973.Find this resource:

Bajari, P., Hong, H., Park, M., & Town, R. (2011). Regression discontinuity designs with an endogenous forcing variable and an application to contracting in health care (NBER Working Paper No. 17643). National Bureau of Economic Research.

Barseghyan, L., Molinari, F., O’Donoghue, T., & Teitelbaum, J. C. (2018). Estimating risk preferences in the field. Journal of Economic Literature.Find this resource:

Barsky, R. B., Juster, F. T., Kimball, M. S., & Shapiro, M. D. (1997). Preference parameters and behavioral heterogeneity. An experimental approach in the health and retirement study. The Quarterly Journal of Economics, 112(2), 537–579.Find this resource:

Bates, D. W., & Gawande, A. A. (2000). The impact of the Internet on quality measurement. Health Affairs, 19(6), 104–114.Find this resource:

Besedeš, T., Deck, C., Sarangi, S., & Shor, M. (2012). Age effects and heuristics in decision making. Review of Economics and Statistics, 94(2), 580–595.Find this resource:

Bhargava, S., Loewenstein, G., & Sydnor, J. (2017). Choose to lose: Health plan choices from a menu with dominated options. Quarterly Journal of Economics, 132(3), 1319–1372.Find this resource:

Blondel, S., Loheac, Y., & Rinaudo, S. (2007). Rationality and drug use: An experimental approach. Journal of Health Economics, 26(3), 643–658.Find this resource:

Boes, S., & Gerfin, M. (2016). Does full insurance increase the demand for health care? Health Economics, 25(11), 1483–1496.Find this resource:

Brunello, G., Fort, M., Schneeweis, N., & Winter-Ebmer, R. (2016). The causal effect of education on health: What is the role of health behaviors? Health Economics, 25(3), 314–336.Find this resource:

Buchmueller, T., Fiebig, D., Jones, G., & Savage, E. (2013). Preference heterogeneity and selection in private health insurance: The case of Australia. Journal of Health Economics, 32(5), 757–767.Find this resource:

Bünnings, C. (2016). Does new health information affect health behaviour? The effect of health events on smoking cessation. Applied Economics, 49(10), 987–1000.Find this resource:

Bünnings, C., Schmitz, H., Tauchmann, H., & Ziebarth, N. (2018). The role of prices relative to supplemental benefits and service quality in health plan choice. Journal of Risk and Insurance.Find this resource:

Cabral, M. (2017). Claim timing and ex post adverse selection. The Review of Economic Studies, 84(1), 1–44.Find this resource:

Card, D., Dobkin, C., & Maestas, N. (2008). The impact of nearly universal insurance coverage on health care utilization: Evidence from Medicare. American Economic Review, 98(5), 2242–2258.Find this resource:

Card, D., Dobkin, C., & Maestas, N. (2009). Does Medicare save lives? Quarterly Journal of Economics, 124(2), 597–636.Find this resource:

Cawley, J., & Philipson, T. (1999). An empirical examination of information barriers to trade in insurance. American Economic Review, 89(4), 827–846.Find this resource:

Chandra, A., Gruber, J., & McKnight, R. (2010). Patient cost-sharing, hospitalization offsets, and the design of optimal health insurance for the elderly. American Economic Review, 100(1), 193–213.Find this resource:

Charness, G., Gneezy, U., & Imas, A. (2013). Experimental methods: Eliciting risk preferences. Journal of Economic Behavior and Organization, 87, 43–51.Find this resource:

Chetty, R. (2009). Sufficient statistics for welfare analysis: A bridge between structural and reduced-form methods. Annual Review of Economics, 1, 451–488.Find this resource:

Chiappori, P.-A., Durand, F., & Geoffard, P.-Y. (1998). Moral hazard and the demand for physician services: First lessons from a French natural experiment. European Economic Review, 42(3–5), 499–511.Find this resource:

Chiappori, P.-A., & Salanie, B. (2001). Testing for asymmetric information in insurance markets. Journal of Political Economy, 108(1), 56–78.Find this resource:

Clement, J. P., Bazzoli, G. J., & Zhao, M. (2012). Nursing home price and quality responses to publicly reported quality information. Health Services Research, 47(1 Pt1), 86–105.Find this resource:

Cockx, B., & Brasseur, C. (2003). The demand for physician services: Evidence from a natural experiment. Journal of Health Economics, 22(6), 881–913.Find this resource:

Currie, J., & Moretti, E. (2003). Mother’s education and the intergenerational transmission of human capital: Evidence from college openings. The Quarterly Journal of Economics, 118(4), 1495–1532.Find this resource:

Cutler, D. M., Finkelstein, A., & McGarry, K. (2008). Preference heterogeneity and insurance markets: Explaining a puzzle of insurance. American Economic Review, 98(2), 157–162.Find this resource:

Cutler, D. M., & Glaeser, E. (2005). What explains differences in smoking, drinking, and other health-related behaviors? American Economic Review, 95(2), 238–242.Find this resource:

Cutler, D. M., Huckman, R. S., & Landrum, M. B. (2004). The role of information in medical markets: An analysis of publicly reported outcomes in cardiac surgery. American Economic Review, 94(2), 342–346.Find this resource:

Cutler, D. M., & Lleras-Muney, A. (2010). Understanding differences in health behaviors by education. Journal of Health Economics, 29(1), 1–28.Find this resource:

Cutler, D. M., & Zeckhauser, R. J. (2000). The anatomy of health insurance. In A. Culyer & J. P. Newhouse (Eds.), Handbook of health economics (pp. 563–643). Amsterdam: Elsevier.Find this resource:

Dalton, C. (2014). Estimating demand elasticities using nonlinear pricing. International Journal of Industrial Organization, 37, 178–191.Find this resource:

Dalton, C., Gowrisankaran, G., & Town, R. (2015). Myopic and complex dynamic incentives: Evidence from Medicare Part D (NBER Working Paper No. 21104). National Bureau of Economic Research.

Decker, S., & Schmitz, H. (2016). Health shocks and risk aversion. Journal of Health Economics, 50, 156–170.Find this resource:

Dohmen, T., Falk, A., Huffman, D., Sunde, U., Schupp, J., & Wagner, G. G. (2011). Individual risk attitudes: Measurement, determinants, and behavioral consequences. Journal of the European Economic Association, 9(3), 522–550.Find this resource:

Doiron, D., Jones, G., & Savage, E. (2008). Healthy, wealthy and insured? The role of self-assessed health in the demand for private health insurance. Health Economics, 17(3), 317–334.Find this resource:

Downs, J. S., Loewenstein, G., & Wisdom, J. (2009). Strategies for promoting healthier food choices. American Economic Review, 99(2), 159–164.Find this resource:

Dranove, D., Kessler, D., McClellan, M., & Satterthwaite, M. (2003). Is more information better? The effects of “report cards” on health care providers. The Journal of Political Economy, 111(3), 555–588.Find this resource:

Dupas, P. (2011). Do teenagers respond to HIV risk information? Evidence from a field experiment in Kenya. American Economic Journal: Applied Economics, 3(1), 1–34.Find this resource:

Einav, L., Finkelstein, A., & Schrimpf, P. (2015). The response of drug expenditure to nonlinear contract design: Evidence from Medicare Part D. Quarterly Journal of Economics, 130(2), 841–899.Find this resource:

Epstein, A. J. (2010). Effects of report cards on referral patterns to cardiac surgeons. Journal of Health Economics, 29(5), 718–731.Find this resource:

Ericson, K. M., & Starc, A. (2012). Heuristics and heterogeneity in health insurance exchanges: Evidence from the Massachusetts connector. American Economic Review, 102(3), 493–497.Find this resource:

Fang, H., Keane, M. P., & Silverman, D. (2008). Sources of advantageous selection: Evidence from the Medigap insurance market. Journal of Political Economy, 116(2), 303–350.Find this resource:

Finkelstein, A., & McGarry, K. (2006). Multiple dimensions of private information: Evidence from the long-term care insurance market. American Economic Review, 96(4), 938–958.Find this resource:

Finkelstein, A., & Poterba, J. (2004). Adverse selection in insurance markets: Policyholder evidence from the U.K. annuity market. Journal of Political Economy, 112(1), 183–208.Find this resource:

Finkelstein, A., & Poterba, J. (2014). Testing for asymmetric information using “unused observables” in insurance markets: Evidence from the U.K. annuity market. Journal of Risk and Insurance, 81(4), 709–734.Find this resource:

Finkelstein, A., Taubman, S., Wright, B., Bernstein, M., Gruber, J., Newhouse, J. P., . . . Baicker K. (2012). The Oregon health insurance experiment: Evidence from the first year. The Quarterly Journal of Economics, 127(3), 1057–1106.Find this resource:

Fossen, F., & Glocker, D. (2017). Stated and revealed heterogeneous risk preferences in educational choice. European Economic Review, 97, 1–25.Find this resource:

Frank, R. G., & Lamiraud, K. (2009). Choice, price competition and complexity in markets for health insurance. Journal of Economic Behavior & Organization, 71(2), 550–562.Find this resource:

Gerfin, M., Kaiser, B., & Schmid, C. (2015). Health care demand in the presence of discrete price changes. Health Economics, 24(9), 1164–1177.Find this resource:

Gerking, S., & Khaddaria, R. (2012). Perceptions of health risk and smoking decisions of young people. Health Economics, 21(7), 865–877.Find this resource:

Gloede, O., Menkhoff, L., & Waibel, H. (2015). Shocks, individual risk attitude, and vulnerability to poverty among rural households in Thailand and Vietnam. World Development, 71, 54–78.Find this resource:

Grabowski, D. C., & Town, R. J. (2011). Does information matter? Competition, quality, and the impact of nursing home report cards. Health Services Research, 46(6 Pt. 1), 1698–1719.Find this resource:

Grimard, F., & Parent, D. (2007). Education and smoking: Were Vietnam war draft avoiders also more likely to avoid smoking? Journal of Health Economics, 26(5), 896–926.Find this resource:

Hammar, H., & Johansson-Stenman, O. (2004). The value of risk-free cigarettes—Do smokers underestimate the risk? Health Economics, 13(1), 59–71.Find this resource:

Harrison, G. W., Lau, M. I., & Rutstrom, E. E. (2010). Individual discount rates and smoking: Evidence from a field experiment in Denmark. Journal of Health Economics, 29(5), 708–717.Find this resource:

Heiss, F., Leive, A., McFadden, D., & Winter, J. (2013). Plan selection in Medicare Part D: Evidence from administrative data. Journal of Health Economics, 32(6), 1325–1344.Find this resource:

Herr, A., Nguyen, T.-V., & Schmitz, H. (2016). Public reporting and the quality of care of German nursing homes. Health Policy, 120(10), 1162–1170.Find this resource:

Holt, C A., & Laury, S. K. (2002). Risk aversion and incentive effects. American Economic Review, 92(5), 1644–1655.Find this resource:

Javitt, J. C., Rebitzer, J. B., & Reisman, L. (2008). Information technology and medical missteps: Evidence from a randomized trial. Journal of Health Economics, 27(3), 585–602.Find this resource:

Jensen, R., & Lleras-Muney, A. (2012). Does staying in school (and not working) prevent teen smoking and drinking? Journal of Health Economics, 31(4), 644–657.Find this resource:

Kamhöfer, D. A., Schmitz, H., & Westphal, M. (2017). Heterogeneity in marginal nonmonetary returns to higher education. Journal of the European Economic Association.Find this resource:

Kan, K., & Tsai, W.-D. (2004). Obesity and risk knowledge. Journal of Health Economics, 23(5), 907–934.Find this resource:

Keeler, E. B., & Rolph, J. E. (1988). the demand for episodes of treatment in the health insurance experiment. Journal of Health Economics, 7(4), 337–367.Find this resource:

Kemptner, D., Jürges, H., & Reinhold, S. (2011). Changes in compulsory schooling and the causal effect of education on health. Evidence from Germany. Journal of Health Economics, 30(2), 340–354.Find this resource:

Kenkel, D. S. (1991). Health behavior, health knowledge, and schooling. Journal of Political Economy, 99(2), 287–305.Find this resource:

Khwaja, A., Silverman, D., Sloan, F., & Wang, Y. (2009). Are mature smokers misinformed? Journal of Health Economics, 28(2), 385–397.Find this resource:

Kowalski, A. E. (2015). Estimating the tradeoff between risk protection and moral hazard with a nonlinear budget set model of health insurance. International Journal of Industrial Organization, 43, 122–135.Find this resource:

Laschober, M., Maxfield, M., Felt Lisk, S., & Miranda, D. J. (2007). Hospital response to public reporting of quality indicators. Health Care Financing Review, 28(3), 61–76.Find this resource:

Lin, W., & Sloan, F. (2015). Risk perceptions and smoking decisions of adult Chinese men. Journal of Health Economics, 39, 60–73.Find this resource:

Lundborg, P., & Andersson, H. (2008). Gender, risk perceptions, and smoking behavior. Journal of Health Economics, 27(5), 1299–1311.Find this resource:

Manning, W., Newhouse, J., Duan, N., Keeler, E., Leibowitz, A., & Marquis, M. (1987). Health insurance and the demand for medical care: Evidence from a randomized experiment. American Economic Review, 77(3), 251–277.Find this resource:

McWilliams, J. M., Afendulis, C. C., McGuire, T. G., & Landon, B. E. (2011). Complex Medicare Advantage choices may overwhelm seniors—especially those with impaired decision making. Health Affairs, 30(9), 1786–1794.Find this resource:

de Meza, D., & Webb, D. C. (2001). Advantageous selection in insurance markets. RAND Journal of Economics, 32(2), 249–262.Find this resource:

Miller, A. R., & Tucker, C. (2014). Health information exchange, system size and information silos. Journal of Health Economics, 33, 28–42.Find this resource:

Nightingale, F. (1858). Notes on matters affecting the health, efficiency, and hospital administration of the British army: Founded chiefly on the experience of the late war. London: Harrison and Sons.Find this resource:

Nyman, J. A. (2006). The value of health insurance. In A. Jones (Ed.), The Elgar companion to health economics. Cheltenham, UK: Edward Elgar.Find this resource:

Park, J., Konetzka, R. T., & Werner, R. M. (2011). Performing well on nursing home report cards: Does it pay off? Health Services Research, 46(2), 531–554.Find this resource:

Park, J., & Werner, R. M. (2011). Changes in the relationship between nursing home financial performance and quality of care under public reporting. Health Economics, 20(7), 783–801.Find this resource:

van der Pol, M., Hennessy, D., & Manns, B. (2017). The role of time and risk preferences in adherence to physician advice on health behavior change. European Journal of Health Economics, 18, 373–386.Find this resource:

van der Pol, M., & Ruggeri, M. (2008). Is risk attitude outcome specific within the health domain? Journal of Health Economics, 27(3), 706–717.Find this resource:

Pope, D. G. (2009). Reacting to rankings: Evidence from “America’s Best Hospitals.” Journal of Health Economics, 28(6), 1154–1165.Find this resource:

Rothschild, M., & Stiglitz, J. E. (1976). Equilibrium in competitive insurance markets: An essay on the economics of imperfect information. The Quarterly Journal of Economics, 90(4), 630–649.Find this resource:

Ruggeri, M., & van der Pol, M. (2012). Is risk attitude really specific within the health context domain? Further evidence from an Italian survey using probability equivalent technique and face-to-face interviews. Health, Risk & Society, 14, 655–666.Find this resource:

Sahm, C. R. (2012). How much does risk tolerance change? The Quarterly Journal of Finance, 2(4).Find this resource:

Sato, M., & Ohkusa, Y. (2003). The relationship between smoking initiation and time discount factor, risk aversion and information. Applied Economics Letters, 10(5), 287–289.Find this resource:

Schmitz, H. (2011). Direct evidence of risk aversion as a source of advantageous selection in health insurance. Economics Letters, 113(2), 180–182.Find this resource:

Schmitz, H. (2012). More health care utilization with more insurance coverage? Evidence from a latent class model with German data. Applied Economics, 44(34), 4455–4468.Find this resource:

Schmitz, H., & Stroka, M. A. (2014). Do elderly choose nursing homes by quality, price or location? (Ruhr Economic Papers No. 495).

Schmitz, H., & Ziebarth, N. (2017). Does price framing affect the consumer price sensitivity of health plan choice? Journal of Human Resources, 52(1), 88–127.Find this resource:

Schram, A., & Sonnemans, J. (2011). How individuals choose health insurance: An experimental analysis. European Economic Review, 55(6), 799–819.Find this resource:

Schreyögg, J., & Grabka, M. M. (2010). Copayments for ambulatory care in Germany: A natural experiment using a difference-in-difference approach. European Journal of Health Economics, 11, 331–341.Find this resource:

Schurer, S. (2015). Lifecycle patterns in the socioeconomic gradient of risk preferences. Journal of Economic Behavior & Organization, 119, 482–495.Find this resource:

Shigeoka, H. (2014). The effect of patient cost sharing on utilization, health, and risk protection. American Economic Review, 104(7), 2152–2184.Find this resource:

Sinaiko, A. D., Eastman, D., & Rosenthal, M. B. (2012). How report cards on physicians, physician groups, and hospitals can have greater impact on consumer choices. Health Affairs, 31(3), 602–611.Find this resource:

Sinaiko, A. D., & Hirth, R. A. (2011). Consumers, health insurance and dominated choices. Journal of Health Economics, 30(2), 450–457.Find this resource:

Sloan, F., & Platt, A. (2011). Information, risk perceptions, and smoking choices of youth. Journal of Risk and Uncertainty, 42(2), 161–193.Find this resource:

Stigler, G. J., & Becker, G. S. (1977). De gustibus non est disputandum. The American Economic Review, 67(2), 76–90.Find this resource:

Taubman, S. L., Allen, H. L., Wright, B. J., Baicker, K., & Finkelstein, A. N. (2014). Medicaid increases emergency-department use: Evidence from Oregon’s health insurance experiment. Science, 343(6168), 263–268.Find this resource:

Varkevisser, M., van der Geest, S. A., & Schut, F. T. (2012). Do patients choose hospitals with high quality ratings? Empirical evidence from the market for angioplasty in the Netherlands. Journal of Health Economics, 31(2), 371–378.Find this resource:

Vera-Hernandez, M. (2003). Structural estimation of a principal-agent model: moral hazard in medical insurance. RAND Journal of Economics, 34(2003), 670–693.Find this resource:

Wagner, T. H., Hu, T.-w., & Hibbard, J. H. (2001). The demand for consumer health information. Journal of Health Economics, 20(6), 1059–1075.Find this resource:

deWalque, D. (2007). Does education affect smoking behaviors? Evidence using the Vietnam draft as an instrument for college education. Journal of Health Economics, 26(5), 877–895.Find this resource:

Wang, J., Hockenberry, J., Chou, S.-Y., & Yang, M. (2011). Do bad report cards have consequences? Impacts of publicly reported provider quality information on the CABG market in Pennsylvania. Journal of Health Economics, 30(2), 392–407.Find this resource:

Werner, R. M., Konetzka, R. T., & Kruse, G. B. (2009). Impact of public reporting on unreported quality of care. Health Services Research, 44(2 Pt. 1), 379–398.Find this resource:

Werner, R. M., Norton, E. C., Konetzka, R. T., & Polsky, D. (2012). Do consumers respond to publicly reported quality information? Evidence from nursing homes. Journal of Health Economics, 31(1), 50–61.Find this resource:

Wisdom, J., Downs, J. S., & Loewenstein, G. (2010). Promoting healthy choices. Information versus convenience. American Economic Journal: Applied Economics, 2(2), 164–178.Find this resource:

Ziebarth, N. R. (2010). Estimating price elasticities of convalescent care programmes. Economic Journal, 120(545), 816–844.Find this resource:

Zweifel, P., Breyer, F., & Kifmann, M. (2009). Health economics (2nd ed.). New York: Springer.Find this resource:

Zweifel, P., & Manning, W. G. (2000). Moral hazard and consumer incentives in health care. In A. J. Culyer & J. P. Newhouse (Eds.), Handbook of health economics (pp. 409–459). Amsterdam: Elsevier.Find this resource:

Notes:

(1.) As common throughout this article we, with some exceptions, focus on more recent literature. Much of the earlier literature on this topic is surveyed by Cutler and Zeckhauser (2000).

(2.) Parts of this section draw from an unpublished manuscript by H. Schmitz called “Risk Aversion, Health Behaviour, and Adverse Selection in the German Market for Private Supplementary Health Insurance.”

(3.) Or, in a separating equilibrium, bad risks buy full-cover insurance while good risks buy less-than-full-cover insurance—that is, insurance with higher deductibles.