Beware of Weaponized Statistics
A couple of weeks ago, my colleague Dr. Don Mathews published a From the Murphy Center column on the deceptive qualities of percentages when it comes to understanding inflation. In his column, he noted that the same percentage change will mean a small numerical change if the initial value is small and a large numerical change if the initial value is large. Dr. Mathews encouraged readers to look at the initial value to avoid being “deceived” by a percentage. I hope to build on his sage advice.
We are bombarded by statistics all the time on the economy, crime, public health, and other social problems. Americans tend to privilege quantitative evidence (statistics) as objective facts about social phenomena. However, all statistics are social constructions; statistics are the products of organizations and individuals who create them.
In his book, Damned Lies and Statistics, Dr. Joel Best asserts that activists, reporters, experts, officials, and private organizations have their own reasons to promote particular social issues. Best coins the term “problem promoters” to emphasize that people and organizations deliberately construct social problems. Problem promoters often choose one statistic over others to shape our opinions about social phenomena. We must remain vigilant not to be duped.
Weaponized statistics often take the form of percentages, absolute numbers, or comparisons. A problem promoter often chooses a statistic to “prove” their point and sway our opinions on public policy. Consider how statistics on US overdose deaths are weaponized to shape our perspectives on this social problem.
A May 18, 2023 press release from the National Center for Health Statistics offers insight into US overdose deaths: “Provisional data show that the reported number of drug overdose deaths occurring in the United States decreased by 2% from the 12 months ending in December 2021 to the 12 months ending in December 2022.” This indicates that we may have turned a corner in the opioid crisis. Alternatively, if one accounts for delays in reporting, the NCHS projects a 0.5% increase in overdose deaths for this same time period. This statistic leads one to conclude that the opioid epidemic is worsening. Without knowing the initial value, it’s hard to determine whether any significant change has occurred.
Absolute numbers are often used to highlight the severity of phenomena. After adjustments for delayed reporting, the projected number of drug overdose deaths in the US in 2022 was 109,680, according to the NCHS press release referenced earlier.
Additionally, comparisons are used to sway our opinions. A July 12, 2023 press release from the White House states, “As the Biden-Harris Administration works to fully implement President Biden’s National Drug Control Strategy, US is seeing continued progress after a full year of flattening overdose deaths, halting [a] period of rapid increases from 2019-2021.” This statistic celebrates plateaued overdose deaths in 2022-2023 compared to rising overdose deaths in 2019-2021. Comparisons can be based on quantitative or qualitative data.
Organizations and individuals with a particular political agenda weaponize one statistic over others to shape our attitudes about overdose deaths in the US. These statistics color our attitudes about drug use, the opioid crisis, and related public policy.
We experienced statistics weaponization amid the COVID-19 pandemic. Medical experts, government officials, and media personalities shared statistic on COVID-19 deaths, survival rates, and vaccine efficacy data to shape individual behavior and public policy.
Reflecting on weaponized statistics can lead one to nihilism or cynicism. Alternatively, I challenge readers to be critical consumers of all quantitative or qualitative evidence. Be mindful about how problem promoters present their statistics as percentages, absolute numbers, or comparisons to influence uncritical readers.
Dr. Joel Best offers some additional advice. He challenges us to ask three questions of any statistic: Who created this statistic? Why was this statistic created? How was this statistic created? These questions help us to sort out good statistics from bad statistics.
Roscoe Scarborough, Ph.D. is chair of the Department of Social Sciences and associate professor of sociology at College of Coastal Georgia. He is an associate scholar at the Reg Murphy Center for Economic and Policy Studies. He can be reached by email at rscarborough@ccga.edu.