Communicating Uncertainty Is Part of Science

By Toby Shu

What should scientific institutions do to maintain their public trust while still contributing to good policy? … “Scientists 90% sure lockdowns a good idea” does not make a good headline. 

Science is defined by doubt and ignorance. The scientific method begins with a question, a lack of knowledge, or admitted ignorance, then forces us to form a prediction that will likely be proved wrong. A theory about the world is discovered through many observations, but a single outlier fact could cast the theory into doubt. History reflects that scientific paradigms have shifted time and time again, each old view being cast into the pile of abandoned theories and pseudoscience. 

Science is nonetheless called upon to deliver facts and theories about the world that can be enacted into policy to shape the world. Even though science is seen as fact-based certainty, it is actually a dynamic process of unsettled knowledge. Thus, we must ask how scientists can shape policy while understanding and communicating their doubt and ignorance. In order for scientists to maintain their standing as trusted experts, they cannot excessively emphasize their uncertainty, nor can they overconfidently state their claims. To do either would result in long-term loss of trust in scientific institutions.

Let’s take a look at the scientific response to COVID-19, as represented by representatives of the medical and scientific establishment, like the CDC and Dr. Anthony Fauci. At the beginning of the pandemic, scientific institutions recognized the extreme danger of the novel coronavirus. Specifically, they recognized the threat posed to personnel, medical resources, and the public at large. In order to combat these dangers, quarantines and shutdowns were recommended in order to “flatten the curve,” spreading out cases over time to ensure services would not be overwhelmed. However, the information given was often more uncertain than was communicated to the public, partially because some wanted to ensure drastic and necessary action would be taken in order to save lives and partially because uncertainty can get deemphasized as information travels along the grapevine and to other people. “Scientists 90% sure lockdowns a good idea” does not make a good headline. 

This failure to communicate uncertainty in the face of a novel threat led to distrust. For instance, there was initially conflicting guidance on whether masks should be worn by the public. Many health experts wanted masks to go to at-risk medical personnel at the front lines fighting the pandemic, and therefore suggested that masks were inefficacious for the public. This guidance changed within a month and a half, with health experts coming to the new conclusion that masks were beneficial for the public and preventing the disease from spreading. This guidance went on to be enforced by mask mandates. This rapid change in recommended behavior combined with governmental enforcement led to backlash by skeptics, who saw this policy change as scientifically questionable and a means of authoritarian control. This is one of a number of changes in understanding about COVID-19 that led to shifting science communication and policy from scientific authorities. None of these changes had the effect of increasing confidence in scientific institutions.

Of course, not all lost trust in science due to COVID communication was the fault of scientists. Many people were predisposed to question scientific authorities, and used shifting understandings as a source for confirmation bias about the scientific establishment. Misinformation abounded on social media, and other bad actors sought to exacerbate the chaos caused by lockdowns. However, by failing to discuss uncertainty, scientific institutions contributed to the loss of public confidence as well. For instance, little to no uncertainty was communicated about scientific positions on COVID’s surface transmissibility, the lab leak theory, and children’s vulnerability to COVID. When findings changed and prior scientific understandings were shown to be not entirely correct, public confidence was lost in scientific institutions. When scientific institutions fail to discuss the uncertainty in their conclusions, it is significantly easier to understand them as not driven by the science—especially when they exist in a media sphere surrounded by other disinformation. 

Thus, what should scientific institutions do to maintain their public trust while still contributing to good policy? They should underline their uncertainty and prevent it from being removed from the conversation. They should acknowledge mistakes made openly and explain why they were made. They should give the facts as they understand them to policymakers, and allow the policymakers to craft appropriate policy. 

By being upfront and accurate about uncertainty, scientists are honest and future changes in understanding lose power to lessen trust. The domain of science is understanding the world and explaining cause and effect. Scientists are not better situated than policymakers and elected representatives to make judgments about questions of moral value for the public, such as how to best weigh freedom against preventing disease spread or how triage should look.

To maintain credibility, scientific institutions should separate themselves from the policy process and specific political outcomes, instead focusing on their role in providing credible expert information. Though science feeds into policy, scientific institutions’ primary role is to provide sound science. Science understands the natural world and how actions will affect it, and is separate from ethics and democratic will. Thus, scientific institutions should communicate objective data, diagnosing phenomena and explaining causes and effects, leaving policymaking to elected representatives.