By Phil La Duke
Statistics are tricky. Because they are expressions of probability one can be mislead by statistic, as the old saying goes, “statistics lie, and liars use statistics”. In the world of safety perhaps the most widely quoted, pervasive, and well…just plain wrong, statistic is that 95% of all injuries are caused by unsafe behaviors. It’s a tidy and convenient statistic that is cherished by both Operations and Safety professionals. People like this statistic, chiefly because it puts the onus on the worker for staying safe. It holds, that 95% of all injured workers are to blame for their injuries (or at a minimum, another worker’s behavior is responsible for their injury.) It lets both companies and safety professionals off the hook—they can’t be held culpable for workers who refuse to work safely.
Unfortunately, there are serious issues with this statistic. In 2001, noted safety theorist, Fred A. Manuele, published “Heinrich Revisited: Truisms or Myths” (Herbert William Heinrich was an insurance professional who, in the 1930s investigated the causes of workplace injuries). Manuele set out to study the origins of this statistics and basically to determine the veracity of this statistic. The book is a great read, and all those ninnies who are out there proclaim these statistics as Gospel should read it. I won’t take a great deal of time dissecting it (like I said, buy the man’s book), but I think there are some observations of which safety professionals should be aware (the following are :
- The Research is Not Replicable Heinrich either didn’t keep records of his research, didn’t keep it, or it has been lost. Some safety professionals (not necessarily Manuele) interpret this as indicative that Heinrich made up his findings. I personally doubt that Heinrich faked his work or made up his findings, but under the scientific method, one is expected to keep records of one’s research so that your professional peers can review it and critic it before publication. So if there is no research there can be no peer review and the work is not accepted as valid. So Heinrich’s work was the 1930’s version of a blog, an opinion piece that (like every good urban legend) made sense and sounded right to people eager to believe it.
- That Was Then. This is Now. Manuele points out that Heinrich conducted his research in the 1920’s a scant nine years after the Triangle Shirtwaist Fire that created an uproar that ushered in workplace safety reforms. As Manuele observed, so much has changed in workplace safety that 70-year old research needs to be taken with a grain of salt.
- Psychology Is Not A Cure All. Manuele points out that Heinrich greatly over estimated the role of psychology in accident causation. Heinrich believed that psychology was of “a fundamental of great importance in accident causation” and his love of, and rabid belief in, psychology blinded him to other, potentially more important contributors to injuries. It’s not that Heinrich’s beliefs in psychology were misguided—understanding why people make mistakes, make bad decisions, and take risks is of paramount importance in safety—but the belief that Skinnerian behavior modification could somehow make the workplace safer was misguided. Heinrich can be forgiven for embracing these theories. Freud, Jung, Skinner, and other founding fathers of psychology were still doing important research and publishing findings. But unlike the work of these behavioral scientists, little to no follow up has advanced Heinrich’s theories.
- Many of Heinrich’s Other Conclusions Are Even more Specious. Some of Heinrich’s other conclusions directly conflict with the research and findings of later, more respected management experts like W. Edwards Deming and Peter Drucker. Deming in particular found that root causes of process failures (of which injuries are a symptom) grew out of flaws in management systems. In short, management research from the 1950’s on tend to call Heinrich’s findings into question.
- Like Many Of His Time, Heinrich Believed That Automation Was the Answer To All Of Life’s Problem’s. Walter Ruther believed that “automation will be the salvation of the working man.” Ruther believed that machines would one day do the most dangerous jobs. Heinrich shared this belief, and like many contemporaries believed that “man failure is the heart of the problem and the methods of control must be directed toward man failure.” This belief was the prevalent opinion in business in the 1920’s and 30’s. But the work of Drucker and Deming supported a more enlightened view of industry. In their view, workers possess invaluable information relating to process capability and overall process improvement. As jobs become more and more sophisticated, they typically require workers who are able to think and make sound decisions to keep the process operating.
- Heinrich Was Kind Of a Bigot. Like many of his day, Heinrich believed that people of a certain ancestry or from a given socioeconomic background were intrinsically…well, stupid. In the minds of many, these people were pretty much incapable of learning to do things without hurting themselves. It is easy to see how someone of this belief set could, subconsciously conclude that workers of a given ethnicity would be responsible for 95% of all accidents. It’s worth noting that Heinrich’s research was conducted in the 1930’s, before the Nazi atrocities pretty much killed off the idea of eugenics. World War II forever changed the landscape of behavioral science and theories based on the belief that one race, ethnicity, or socioeconomic background being intrinsically biologically or intellectually superior quickly went out of vogue.
- Most of Heinrich’s Conclusions Were Based on Anecdotal Evidence. From what little we know about Heinrich’s research we know that he asked supervisors what they THOUGHT the causes of injuries were. He didn’t conduct any formal accident investigations nor did he (as far as anyone can tell) talk to the injured workers, many of whom were likely dismissed as unfit for duty.
An entire industry has grown out of research of questionable methods, from a time when industrial processes were largely a simple adaptation of agricultural methods that were conducted by a man who had more in common with the Nazi’s than with Deming. People have trouble abandoning a business philosophy that is paying their kids college tuition. But isn’t about time that we stop giving Heinrich’s theories the same weight and credibility as someone who uses modern scientific methods and produced replicable research that is reviewed and analyzed by his or her peers?