The rational-actor assumption has also dominated many corners of healthcare policy, such as:
The Obama Administration’s current emphasis on pay-for-performance through Accountable Care Organizations
Each of these policies is designed to deliver cost-effective care or to prevent disease by modifying patient behavior. To continue making progress in these areas, however, will require a more complex and complete understanding of human decision making. We need to expand beyond the 19th-Century notions that drive much of our current approach to health administration and policy.
This very endeavor was the focus of a plenary session of the 5th Biennial Conference of the American Society of Health Economists, held on the campus of the University of Southern California and hosted by the USC Schaeffer Center for Health Policy & Economics. Panelists recounted findings from recent research in behavioral economics and social psychology that point to practical advice for hospitals, physicians, insurers, and the growing health information- technology industry.
What these insights share in common is they challenge the two main pillars of rationality as defined by economists: (1) a decisionmaker who strives to maximize his or her own welfare, and (2) a decisionmaker with sufficient information and intelligence to recognize which option best achieves that goal.
In this post, I focus on four examples where the self-interest assumption can lead us astray when designing approaches to healthcare delivery:
1. Self-interest is only one of many important motivations.
“We know that clinicians are often rational, but not always rational.” USC’s Jason Doctor, explains that policies designed to capitalize on self-interest have frequently underperformed. “If you look at pay-for-performance, it hasn't been as effective as we would have hoped. A Cochrane review in 2011 concluded that there is insufficient evidence to support or not support financial incentives. Again, the data are very mixed. A large study in the UK on pay-for -performance showed that large incentives amounting to about 25% of primary care physicians’ pay seemed to work. However, what happened in this study, which was published in the New England Journal of Medicine, is that non-incentivized measures got worse. So, it's not a simple additive equation. Then, there was a recent JAMA paper in 2013, which was one of the first large randomized trials of pay for performance in the US. They found trends but no statistically significant differences between treatment and control groups for patients who were harder to treat.”
2. Helping others often trumps simple self-interest.
Assuming a rational actor, one might try to improve doctors’ handwashing habits by posting signs reminding them of the importance of handwashing (i.e. giving them relevant information) and appealing to their personal self-interest to avoid contagions. However, UCLA’s Dr. Craig Fox recounted results from a 2011 study in Psychological Science that compared the effectiveness of two types of signs – one with a physician-centered message, ‘Hand hygiene prevents you from catching diseases’ and one with a message focused on patient welfare, ‘Hand hygiene prevents patients from catching diseases.’ “You might guess which one was most effective here… Actually, the reminder focusing on the patients’ health was most effective. So, there's a lot of psychology that goes into tuning these messages, and how you arrange the environment can have profound effects on handwashing.”
3. Honoring one’s commitments often trumps simple self-interest.
Humans are social animals frequently preoccupied with maintaining their social status in groups. One way to maintain status is to cultivate a reputation as a consistent, trustworthy partner. To harness this desire in clinical settings, USC’s Professor Daniella Meeker asked physicians to commit to only prescribing antibiotics when appropriate. Inappropriate prescribing for nonbacterial infections is tempting because it often satisfies the patient, but it carries high social costs including antibiotic resistant bacteria, adverse drug reactions, and wasted healthcare resources. Professor Meeker had clinicians display poster-sized commitment letters in their exam rooms, and using electronic health record data, she determined that it reduced inappropriate prescribing by 20%. “If it was applied nationally, and if that effect held up,” according to collaborator Jason Doctor, “it might save around $70 million over 2.6 million inappropriate prescriptions.”
4. Conformity with social norms often trumps utility maximization.
Instead of carefully calculating which option is optimal, we often rely upon a much simpler heuristic—do what everyone else is doing. Based on empirical research, social psychologists conclude that “descriptive norms” – our beliefs about what is most common – have a stronger influence on our own behavior than “injunctive norms” – our beliefs about what is most socially approved. The interactions between descriptive and injunctive norms can have important ramifications for public health. For example, a recent study in the Journal of Applied Social Psychology finds that injunctive norms urging people to adopt a healthy diet can actually backfire for people who know that their friends and family actually eat poorly (a negative descriptive norm). This result mirrors prior evidence that “Do not litter” signs actually encourage littering when placed in heavily littered areas. In this case, the main effect of the injunctive message is to remind potential litterers that most people do not follow the rules; the dominant descriptive norm prevails. “Everyone litters so why shouldn’t I?” If hand-washing signs induce the same type of cognitive response in healthcare workers, they might do more harm than good in the hospitals that seemingly need them most. Hippocrates would not be pleased.
Most aspects of healthcare administration – from managing a hospital to writing the nation’s health laws – involve nudging human behavior in desired directions. To succeed, we need a more modern and holistic view of human motivation. We need to wise up about rational choice.