Are humans unrealistically hopeful about the future, or just pretending to be?
Whether you know it or not, you’re probably an optimist. That’s what most experts in the social sciences contend, based on a mountain of studies showing that people consistently report being at less risk than others of experiencing negative life events. The so-called “optimism bias” persists even after people have been given true information about risk, research finds. For their part, many neuroscientists argue that our brains are engineered for optimism, which if true would certainly explain the durability of the bias. Popular media’s embrace of this research has recently produced a bestselling book and a spate of articles blaming unrealistic optimism for everything from stock market bubbles to the election victory of Donald Trump.
Since the optimism bias was first discovered in a seminal 1980 study, researchers have been trying to figure out how it works. Some hypothesise that the bias is more sensitive to certain types of negative events, e.g. those that are more severe, frequent or aligned with social stereotypes. In particular, unhappy outcomes that are presumed to be controllable (and hence, avoidable) are associated with high levels of unrealistic optimism. In other words, the more control you have over the likelihood of something bad happening to you, the more you’ll deny your susceptibility to it. Since lung cancer risk is increased by smoking, a voluntary activity, it should automatically inspire more unrealistic optimism than risk for cancers that appear to strike more randomly. Similarly, since risk for sexually transmitted diseases (STD) can be avoided by using appropriate protection, people demonstrate greater unrealistic optimism for STDs than for diseases that cannot be so ‘easily’ avoided (e.g. airborne viruses).
Now, however, imagine a closet smoker participating in such a study. It seems likely that his or her irrationally optimistic responses would be motivated not by controllability per se, but more precisely by embarrassment to admit his or her elevated risk stemming from a totally needless habit. In the case of STDs, people might be reluctant to admit perceived susceptibility for fear of what that would reveal about their private lives. If this pattern applies more widely, there could be a good deal of pretence baked into the optimism bias, a possibility totally neglected in previous research. The predictors identified in past studies could largely be proxies for a correlated cause: embarrassment.
To fill in the gaps, my co-authors (Stefano Puntoni, Gabriele Paolacci, and Maarten Vissers of Rotterdam School of Management) and I revisited data from 28 past studies on optimism bias, encompassing 8,826 participants in all. Our article in Organizational Behavior and Human Decision Processes posits that much of the human response usually ascribed to optimism bias is actually a smokescreen to ward off social disapproval. If so, the core of our personalities may have more to do with caution than cockeyed optimism.
The studies in question covered a total of 79 unique life events, ranging from relatively minor misfortunes (developing arthritis, breaking a bone, etc.) to more serious outcomes (e.g. experiencing vision loss, getting fired). We convened a group of 309 undergraduate students to rate each event’s social undesirability, controllability, severity, frequency and stereotype salience. Using hierarchical linear modelling, we mapped the students’ ratings against the results across all the studies to trace out predictive patterns. Social undesirability was by far the most powerful predictor of comparative optimism, which implies that past studies that omitted this critical characteristic may not have the full story.
We sought to replicate and deepen our findings in four subsequent experiments. In perhaps the most pertinent of these, we attempted to manipulate the optimism bias by providing various levels of anonymity reassurance prior to administering a survey about negative life events. Some participants were asked to sign an enlarged, more elaborate consent form emphasising confidentiality while others received either a standard consent form or one with no mention of confidentiality. The more insistently we promised to protect respondents’ anonymity, the less their optimistic responses were inflated by socially desirable responding. The results confirmed our hypothesis that at least part of the optimism bias is attributable to people’s unwillingness to expose their susceptibilities.
In another experiment, we asked undergraduate women to state their own and their female peers’ likelihood of both falling pregnant and having a car accident within the next year, as well as within the next ten years. The time variation was essential, as pregnancy while one is still a student carries heavy social stigma; pregnancy later in life doesn’t, nor do traffic accidents at any age. As hypothesised, respondents reported their own risk as much lower than others’ only during the near-term pregnancy condition. Again, we can conclude that the so-called “optimism bias” may reflect aversion to social embarrassment more than genuine optimism.
Optimism bias in business
Some scientists believe that by making awareness of mortality easier to bear, optimism played a key role in human beings’ evolutionary development. In today’s Darwinian business world, a show of optimism can be just as advantageous as the real thing, whereas an honest assessment of doubts and vulnerabilities could leave one dangerously exposed.
Even in the B-school classrooms where I’ve taught, the default tendency is toward soaring optimism. Over the last eight years, I’ve guided more than 1,000 MBA and EMBA students through computer marketing simulations and I always ask participants to forecast the performance of their fictitious companies about three virtual years into the future. Consistently, I’ve noticed that the students’ projections overshoot their actual scores by about 150 percent.
Managers may find that feigning optimism inspires loyalty from employees and faith from higher-ups. Stakeholders who believe these cheerful predictions may be in for a rude awakening, however. Our findings suggest that if you want to filter out the noise, you must provide an almost ostentatious level of anonymity protection. In an atmosphere of maximally reinforced reassurance, you have the best chance of banishing the rose-coloured glasses.
Similarly, public-service or advertising campaigns designed to erase complacency about taboo subjects may not be as effective as previously thought, since the complacency is only on the surface. Deep down, people know they’re at risk but are wary of admitting it. A better approach would be to try to wipe out the social stigma so that people feel relaxed enough to air their secret fears, and act upon them. Replace “Yes, it CAN happen to you” with “It could happen to anyone.”
In short, though there does seem to be some amount of reflexive optimism built into our psychological processes, our research implies that decision-making needn’t always be addled by it. It’s ironic that humanity’s prospects look quite a bit more optimistic if the “optimism bias” is largely illusory.
Steven Sweldens has been affiliated with INSEAD since 2009, as an Assistant Professor of Marketing (2009-2015) and as a Distinguished Research Fellow (2015-present). Since 2014, Steven is also an Associate Professor of Marketing at RSM Erasmus University.