Abstract
Dysfunctional health behavior is a contemporary challenge, exemplified by the increasingly significant portion of health problems stemming from people’s own behavior and decision making. The challenge not only includes the direct consequences of unhealthy behavioral patterns but also their origins and the creation of policies that effectively decrease their frequency. A framework rooted in behavioral economics identifies the processes and mechanisms underlying poor health. Two behavioral economic processes, economic demand and delay discounting, are discussed in detail. Through continued development, this behavioral economic framework can guide improved outcomes in treatment and policies related to dysfunctional health behavior. Approaches are evolving to alter demand and discounting. Current and prospective policies aimed at decreasing unhealthy behavior may profit from such research.
Tweet
Many diseases emerge from similar patterns of unhealthy decision making. Behavioral economics offers policy insights.
Key Points
Medical conditions that contribute to excess morbidity and mortality are increasingly the result of human behavior and decision making.
Behavioral economics—combining principles from psychology, cognitive science, and economics—provides a framework to identify principles that undergird human behavior in general and dysfunctional behavior in particular.
The propensity for unhealthy behaviors can be measured by two independent processes, behavioral economic demand and delay discounting.
The processes of demand and delay discounting provide a perspective on potential policy interventions and allow testable hypotheses to develop more effective policy options.
The most effective interventions to address dysfunctional behavior must address high behavioral economic demand and steep discounting of future rewards in the context of socioeconomic status and quality of education.
Introduction
A significant portion of contemporary health problems result from what people choose to do or not do, such as the health challenges associated with addiction, obesity, and medical non-compliance. These behavioral bases of poor health will only increase in the future and will play an increasingly larger role in science, health, health care, and policy. This article explicates the problems of unhealthy behaviors and complications involved in designing policies, identifies some known sources of those behaviors, and reviews the measures and relevance of two behavioral economic processes, demand and delay discounting, as underlying mechanisms of poor health. Understanding the behavioral processes that undergird the health problems so prevalent in our society can guide improved outcomes in addressing these challenges.
Statement of the Problem
Dysfunctional Health Behavior
Over the last two decades, interest in preventing disability and death through changes in health-related behaviors has risen greatly. Much of this interest was motivated by the change in disease patterns from infectious (before 1940) to chronic diseases and cancer (afterward) as leading causes of death, combined with the data associating health behaviors to increased risk of morbidity and mortality. For example, since the 1960s, dysfunctional health behavior, such as cigarette smoking and overeating, has played a critical role in the overall trend in the mortality rate for all leading causes of death in the United States (i.e., heart disease, cancer, stroke, unintentional injuries, obstructive pulmonary disease, and diabetes). However, reductions in heart disease, stroke, and diabetes resulted from improvement in hypertension and hyperlipidemia control, smoking cessation, and advances in treatment (Ford et al., 2007; Lackland et al., 2014; Young, Capewell, Ford, & Critchley, 2010). Rates of death due to cancer have also declined since the 1990s, as a result of tobacco-control efforts in addition to advances in screening, early detection, and treatment (Edwards et al., 2014; Jemal et al., 2008).
By the 1990s, the most common behavioral contributors to mortality (alcohol, tobacco, and illicit drug use; motor vehicles; firearms; diet and activity patterns; and sexual behavior) caused almost half of the deaths in the United States, responsible for nearly 1 million deaths in 1992 alone (McGinnis & Foege, 1993). Furthermore, as medical treatment for more behaviorally independent diseases has improved, the relative share of disease burden due to dysfunctional health behaviors has risen. In 2005, the leading risk factors for adult mortality were tobacco smoking and high blood pressure (often a result of poor diet and sedentary lifestyle choices), which were responsible for approximately one in five and one in six deaths, respectively. In addition, physical inactivity, obesity, and high blood glucose each caused about one in 10 deaths (Danaei et al., 2009). Despite progress in the reduction of mortality over the past six decades, major behavior-related cardiovascular risk factors (e.g., elevated cholesterol levels, diabetes, hypertension, obesity, and smoking) accounted for about 50% of cardiovascular deaths in adults aged 45 to 79 years in 2009-2010 (Patel, Winkel, Ali, Narayan, & Mehta, 2015), and about 3.5% of all cancer deaths (~19,500 people) were caused by alcohol consumption (Nelson et al., 2013).
Policy options for dysfunctional health behavior
Despite behavioral research fundamentally transforming the field of economics by placing it on a firm experimental basis (Camerer, Loewenstein, & Rabin, 2011; Gintis, 2007), traditional economic models remain the benchmark for applications to interventions and policies. These policies have been effective in reducing certain dysfunctional health behaviors, for example, tobacco use. However, many attempts have shown little or stunted effectiveness, while causing other detrimental socioeconomic problems (see Chetty, 2015; Currie, 1994; Felbab-Brown, 2008; Madrian, 2014; Werb et al., 2013). As a result, patterns of unhealthy behavior in America are more frequent and seemingly intractable than ever, and solutions remain elusive (Bickel, Mueller, MacKillop, & Yi, 2016).
The mixed effectiveness of many past and current policies may be due to the lack of understanding of the contexts, contingencies, and controlling variables that drive dysfunctional health behavior (Bickel & Marsch, 2000; Chetty, 2015; Madrian, 2014). Behavioral economics, by combining principles from psychology, cognitive science, and economics, investigates processes contributing to decision making. Thus, as the scientific understanding of behavioral economics increases, so do the tools for policy.
Sources of Unhealthy Behavior
Several unhealthy choices share two common behavioral features: (a) over-valuation of a reinforcer (reward) and (b) excessive preference for an immediately available reinforcer (compared with a delayed one) despite its long-term negative effects. The first contributor to persistent dysfunctional behavior patterns of excessive consumption is the brief, intense reward from unhealthy commodities such as illicit drugs and unhealthy foods. The over-valuation of reinforcers contributes to the high demand for commodities that perpetuate many unhealthy behaviors. For example, substance users will spend much time and money obtaining and using the drug.
Second, many dysfunctional health behaviors choose near-term gratification (e.g., smoking a cigarette to alleviate a craving) at the possible expense of long-term health consequences (e.g., risk of lung cancer or heart disease). Repeatedly selecting short-term reinforcers may occur when an individual has a constricted temporal view (time horizon) that results in near-term reinforcers being excessively valued and long-term consequences being disproportionately discounted. Chronically engaging in unhealthy, but immediately satisfying, behaviors despite long-term negative health consequences accounts, in part, for the high rates of morbidity and mortality associated with dysfunctional health behaviors. The next section discusses these two processes in detail before discussing policy interventions on dysfunctional health behaviors from the behavioral economic lens.
Behavioral Economic Demand and Delay Discounting as Measures of Valuation
Behavioral economic demand (over-valuation) and delay discounting provide a perspective on dysfunctional health behaviors. The focus here will not address whether the underlying assumptions of these models are sound (Bickel, Jarmolowicz, Mueller, & Gatchalian, 2011; for review, see Bickel, Johnson, Koffarnus, MacKillop, & Murphy, 2014). Instead, we will discuss how high behavioral economic demand for certain commodities and excessive devaluation of the future can inform policy interventions to improve health.
Behavioral economic demand
Behavioral economic demand allows experimental examination of the relation between consumption and price at the level of individual subjects. Price (i.e., unit price) is the ratio of price of the reinforcer to the work required to obtain it (Hursh, 1984). By assessing either real or hypothetical measures of consumption across a range of prices, a demand curve may be generated by fitting a non-linear regression model (Hursh & Silberberg, 2008; Koffarnus, Franck, Stein, & Bickel, 2015). The demand curve itself reveals two important parameters that convey an individual’s valuation of a commodity: (a) intensity of demand (i.e., consumption when the commodity is free) and (b) elasticity (i.e., the sensitivity of consumption to price increases). For example, consumption may be manipulated by increasing the price (or the work required) to obtain the reinforcer, or by decreasing the amount of reinforcer delivered. When dealing with multiple commodities, sometimes increasing the price of one results in increased consumption of another (i.e., substitution; Hursh & Silberberg, 2008).
Individual differences in behavioral economic demand are indeed associated with multiple dysfunctional health behaviors. Measures of demand for a given drug are proportional to degree of drug dependence (e.g., cigarettes and alcohol; MacKillop et al., 2010; MacKillop et al., 2008; Murphy, MacKillop, Tidey, Brazil, & Colby, 2011), with those most dependent showing the highest demand. Similarly, in obesity, high demand for food differentiates the obese from the non-obese (Epstein, Dearing, & Roba, 2010; Saelens & Epstein, 1996). Moreover, demand for ultraviolet indoor tanning, a behavior that increases risk for melanoma and multiple carcinomas, differentiates those at greater risk (Reed, Kaplan, Becirevic, Roma, & Hursh, 2016). Substitute reinforcers can reduce demand for highly valued, unhealthy commodities, as demonstrated by research on the demand for drugs given simultaneously available alternative reinforcers (Bickel, DeGrandpre, & Higgins, 1995). Therefore, interventions that decrease demand for unhealthy commodities, and increase substitutes to reduce consumption of unhealthy commodities, may provide therapeutic potential.
Delay discounting
Discounting of delayed rewards decreases value of a reward as a function of the delay to its receipt (Bickel & Marsch, 2001; Kirby, 1997; Madden & Johnson, 2010). For example, receiving US$100 today is typically subjectively more valuable to an individual than receiving US$100 in 5 years. The decline in this subjective value may be empirically measured using a delay discounting task. This task presents a series of preference trials for different amounts of an immediate and a delayed reward (Du, Green, & Myerson, 2002). Repeating this produces a rate of discounting for the individual (Mazur, 1987).
High delay discounting is pervasive in addiction. For example, in cross-sectional studies, individuals meeting clinical criteria for most substance-use disorders, including those for opioids (Madden, Petry, Badger, & Bickel, 1997), cocaine (Coffey, Gudleski, Saladin, & Brady, 2003), and tobacco (Bickel, Odum, & Madden, 1999), show higher delay discounting rates than demographically matched controls (for meta-analyses, see Amlung, Vedelago, Acker, Balodis, & Mackillop, 2016; MacKillop et al., 2011). Likewise, discounting rates are positively correlated with indices of addiction severity (Amlung et al., 2016) and, in prospective studies, high rates of discounting in adolescence predict subsequent initiation of young adult cigarette smoking (Audrain-McGovern et al., 2009).
Delay discounting is also pervasive in other maladaptive health behavior and related outcomes, such as excessive substance use, obesity, and risky sexual behavior (for review, see Bickel & Stein, in press). High rates of delay discounting are a risk factor and perhaps play an etiological role (Bickel, Koffarnus, Moody, & Wilson, 2014) in addiction and other dysfunctional health behavior. Thus, public health policies could operate within the constraints of a narrow temporal view of reward valuation to optimize decision making. Stated differently, proximal rewards are typically more valuable than the same reward presented at a delay; by decreasing the temporal distance of reinforcers, interventions may be more effective.
Empirical Support
Decades of preclinical non-human and human research support demand and delay discounting processes as origins and preservers of dysfunctional health behavior. However, the relevance of these two processes to real-world, unhealthy behavior demands an analysis that crosses multiple levels. The previous sections examined these processes from the view of the individual. The next section reviews evidence from two epidemiological sources to demonstrate the broad applicability of these principles and their potential usefulness in policies to decrease dysfunctional health behavior.
Epidemiological Data
Consider these processes in the behavior of large populations. A now-classic investigation compared heroin consumption in 898 U.S. soldiers stationed at Vietnam and in non-military U.S. residents (Robins, Helzer, & Davis, 1975). In Vietnam, the price of heroin was much lower than in the United States in that the monetary price was lower. Moreover, 85% of the U.S. soldiers were offered heroin, demonstrating its widespread availability, access, and resulting demand. Not only did the prices and demand differ between soldiers in Vietnam and Americans in the United States but also, respectively, 19% and 0.7% became heroin-dependent. In Vietnam, lack of alternative sources of reinforcement promoted preferential focus on immediately available, highly reinforcing heroin (i.e., over-valuation), which may also act to narrow an individual’s temporal view of the future (a steepened discount rate). Moreover, stressful, traumatic, or shocking events shorten one’s temporal view, shifting attention to the present and thereby discounting the future (Bickel, Moody, Quisenberry, Ramey, & Sheffer, 2014; Haushofer & Fehr, 2014; Lawrance, 1991; Mullainathan & Shafir, 2013; Piketty & Saez, 2014). These converging factors, operating through demand and discounting processes, contributed to a 27-fold greater incidence of opioid dependence in the military sample while in Vietnam relative to the domestic incidence.
A second important source of evidence for demand and delay discounting processes at the population level comes from a study by the Global Burden of Disease (GBD) 2013 Obesity Collaboration, which examined the prevalence of overweight and obesity in children and adults from 1980 to 2013 (Ng et al., 2014). The GBD found that, globally, the combined prevalence of overweight and obesity increased by an astounding 27% for adults and 47% for children between 1980 and 2013. One factor is obviously food and its over-consumption. The present analysis suggests that the price of food decreased between those time periods. That is, the number of unhealthy food options increased (e.g., “fast food,” high fructose corn syrup, etc.), portion sizes of unhealthy foods increased (e.g., the super-size), prices for unhealthy foods decreased, prices for healthy foods increased, yielding increased demand for unhealthy foods. Another factor that determines the development of obesity is physical inactivity. Between 1980 and 2013, numerous strong, immediately available sedentary sources of reinforcement (computers, the Internet, and more advanced television systems and channels) were introduced that compete with the delayed beneficial effects of physical activity. Viewed from the framework presented above describing demand and delay discounting as major contributors to dysfunctional health behavior, the reduction in price of unhealthy foods alongside the enhanced discounting of future health brought about by increases in the value of immediately rewarding, sedentary lifestyle behaviors may have contributed to the massive growth in prevalence rates of overweight and obese persons (Barlow, Reeves, McKee, Galea, & Stuckler, 2016; Zimmerman, 2015).
Policy Recommendations
Based on the principles described, effective policies must seek to reduce the burden of dysfunctional health behavior through methods that address high demand for unhealthy commodities and acknowledge limited valuation of temporally distant alternatives to these unhealthy commodities. To this end, policies may increase the functional cost or price of unhealthy behaviors, provide and support alternatives that can serve as substitutes, and, critically, recognize that for individuals with these dysfunctional health behaviors (and a shortened temporal window), any intervention must provide concrete and immediate outcomes. Here, we will discuss two dysfunctional health behaviors, cigarette smoking and illicit drug use, which have been addressed by various policy landscapes, and one rising policy challenge: the management of obesity, with related morbidity and mortality. Attempts to control dysfunctional behavior do not exist in isolation. Just as dysfunctional behaviors are multiply determined, the policies that address them work through multiple processes. Their efficacy can be understood through the extent to which they contact the most relevant components of the decision to engage with unhealthy reinforcers.
Tobacco smoking
One major public health success in the United States has been the reduction of cigarette smoking since 1965 (Fiore et al., 1989). Many effective policies and treatments have operated within the behavioral economic principles of demand and delay discounting. Strategies that have targeted demand include taxation, which directly increases the price of tobacco products, and restrictions on the consumption of tobacco products in certain locations, which increase the non-monetary costs of tobacco consumption (Levy, Chaloupka, & Gitchell, 2004). Because both these cost increases of tobacco consumption are temporally proximate and relatively certain, they operate within a relatively short time frame, supporting the efficacy of these policies among smokers—who discount the future more steeply than non-smokers (Bickel et al., 1999).
Beyond population-level manipulations of price and temporal view, several treatment options support individuals who seek to restrict their own tobacco consumption. For example, nicotine replacement therapies offer lower risk substitutes that alleviate immediate withdrawal and reduce demand for cigarettes (Fiore, Smith, Jorenby, & Baker, 1994; Johnson, Bickel, & Kirshenbaum, 2004). Another treatment for tobacco smoking, contingency management, offers small and immediate incentives for daily decisions to remain abstinent (Higgins, Silverman, & Heil, 2008). Immediate, abstinence-based incentives both provide near-term reinforcers for abstinence (which may be subjectively more valuable than the long-term health benefits of abstinence for people with narrow temporal views) and also increase the unit price of smoking (as it will cost the individual to forego the incentive), which in turn reduces demand for cigarettes. Together, nicotine replacement therapies and contingency management, in addition to other more skill-based therapies such as cognitive behavioral therapy, provide individualized treatment for tobacco use disorders (Cavallo et al., 2007; Oster, Huse, Delea, & Colditz, 1986).
Illicit substance use
In contrast to the successes of tobacco policy, policies to reduce illicit drug use have had less remarkable effects. A cornerstone of current policies designed to disrupt illegal drug use has been the criminalization of their possession and sale, which has had mixed success (Csete et al., 2016; Hart, 2015, 2016; Taylor, Buchanan, & Ayres, 2016). When effective, criminalization likely operates by increasing the price of illegal substances and increasing barriers to consumption (Gallet, 2014; MacCoun, 1993). However, the success of criminalization is limited due to the profitability of black markets (Miron, 2003). Moreover, criminalization, although arguably a high price to pay for substance use, is often a delayed and ambiguous consequence, whether due to the court system or the relatively low probability of being convicted of a crime (although probabilities vary with minority and socioeconomic status; (Banks, 2003). Criminal punishment aims to deter illicit drug use by increasing price, but that price increase is delayed, variable, and inconsistent (Duke & Gross, 2014; Reinarman, 1994). Moreover, criminalization offers no behaviorally effective substitutes, no alternative reinforcement that can operatively compete with drug use (Clear & Frost, 2015; Currie, 1994; Labate, Cavnar, & Rodrigues, 2016).
Alternative and potentially more effective individualized interventions to control illicit substance use have faced barriers to implementation. The community reinforcement approach (Budney & Higgins, 1994)—a therapeutic treatment that offers alternative reinforcers to compete with those offered by substance use—operates within the constructs of behavioral economic demand and substitution, effectively reducing substance use (Meyers & Miller, 2001; Roozen et al., 2004). Furthermore, in the case of opiate use, multiple medicinal substitution therapies (methadone maintenance, buprenorphine treatment) offer lower risk substitutes for opiate use, reducing the immediate costs of abstinence (withdrawal) and allowing contact with other, potentially healthier reinforcers (Buchholz & Saxon, 2016; Mattick, Breen, Kimber, & Davoli, 2014). However, unlike nicotine replacement therapies (available over the counter in all 50 states), most opiate substitution therapies are not readily accessible (Amato et al., 2005). The most effective substitution therapies for individuals who steeply discount the future must instead be conveniently accessible and reduce the delay to the gratification of relief from withdrawal (Greenfield, Brady, Besteman, & De Smet, 1996).
Excessive food consumption
Given that obesity is associated with excessive consumption, policies to control it may develop from similar frameworks as those that have controlled excessive consumption of other commodities, such as tobacco. Effective policy must address factors that promote dysfunctional decision making. Excessive consumption of unhealthy foods has been targeted through taxing or restricting the sale of these commodities (Astrup, 2006; Epstein, Dearing, Roba, & Finkelstein, 2010; McCarthy, 2016; Min, 2013), increasing access to healthier substitutes (Story, Kaphingst, Robinson-O’Brien, & Glanz, 2008) or subsidizing them (Epstein, Dearing, Roba, & Finkelstein, 2010), and offering immediate incentives for consumption of those healthy foods (Horne et al., 2009; see Khan et al., 2009; Tapper, Horne, & Lowe, 2003; Wengreen, Madden, Aguilar, Smits, & Jones, 2013) with varied implementation and outcomes. These policies attempt to increase the immediate price of consumption or offer immediate incentives for choosing healthier substitutes—two strategies that both seek to reduce demand within a framework of high delay discounting.
Conclusion
The future of addressing excess morbidity and mortality increasingly depends on addressing dysfunctional health behaviors. The behavioral economic perspective identifies processes that undergird human behavior in general—and dysfunctional behaviors specifically. As outlined, two processes in particular derived from a behavioral economic understanding are drivers of this dysfunctional behavior. In the case of demand, many commodities that contribute to unhealthy behavior are available at a relatively low cost, leading to excessive consumption. In the case of delay discounting, a shortened temporal view leads to consumption of commodities that are immediately available, without control by long-term consequences of those immediate actions.
These behavioral economic processes not only underlie unhealthy behavior, but also if a particular intervention is effective, it must be operating, in part, through these same processes. This article interprets a range of policies through a behavioral economic framework. This understanding identifies circumstances where these interventions should have little or no effect, as well as the conditions that will enhance their effectiveness. We hope this initial exploration of various procedures that lead to unhealthy behaviors and their correction will further the process of employing behavioral economic demand and discounting of future reinforcers as a core part of policy development.
Footnotes
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.