- A fatal accident in a coal mine decreases the likelihood of another fatal accident by about 85 percent.
- But that effect fades over time and is gone after five years.
- Study author recommends overhauling accident investigations and heightened attention to learning from minor accidents.
Two years ago this month, the Crandall Canyon mine collapsed, killing a total of nine. Has the accident affected the safety of coal mining in the U.S.?
Yes, according to a BYU study of decades of mine disasters. But the research also shows more could be done to preserve the painful lessons learned. One recommendation is steering accident investigations away from finger-pointing and toward uncovering exactly what happened.
“To generalize the results of this study to Crandall Canyon, as tragic and costly as that accident was, Utah mining and this type of mining anywhere in the country will be safer as a result,” said Peter Madsen, assistant professor of organizational behavior and strategy at BYU’s Marriott School of Management. “If the trend identified in the study holds up, we really are on the cusp of a major safety improvement that should happen as a result.”
Here are some of his main findings, to be published in a forthcoming issue of Organization Science, one of the top journals in his field:
- A fatal accident in a given mine decreases the likelihood of another fatal accident in that mine by about 85 percent over the next two years. That protective effect then begins to fade and is back to normal about five years after the accident.
- A minor accident decreases the likelihood of a fatal accident at that mine by 7 percent in the subsequent year, but then that effect disappears.
- A fatal accident at a mine decreases the likelihood of another fatal accident at any mine in the U. S. by about 2.5 to 3 percent.
It’s the first quantitative study on whether organizations learn from disasters, a subject that is usually examined via single-incident case studies.
Madsen analyzed data from the federal Mine Safety and Health Administration on coal mining accidents from 1983 to 2006. That included 938 fatal incidents at the 20,864 mines in the sample.
Although the Crandall Canyon disaster happened after the time frame Madsen studied, applying his overall findings would indicate that its impact on mine safety is still being felt. According to his analysis, two years after a fatal mine accident, other mines in the U.S. are still less likely to suffer another fatal disaster. Three or four years later, the effect strengthens, likely thanks to heightened regulation, before tapering off.
Madsen had expected to find that organizations do learn from disaster, but was taken aback at how soon they forget.
“I was surprised that four or five years after a mine’s fatal accident it was just as unsafe as it had been before the accident,” he said. “I would have expected people to remember a little bit longer.”
Although it might seem obvious that a fatal accident would prompt safety improvements within an organization, that idea is debated by scholars. Some say that disasters are viewed as flukes that couldn’t happen again, or that post-disaster investigations are marred by scapegoating and fail to yield useful conclusions for preventing future mishaps. For example, some researchers question whether NASA learned from the space shuttle Challenger explosion, pointing out the subsequent loss of the Columbia.
Madsen studies how organizations learn from disaster, having contributed to studies of the Columbia disaster as a graduate student at UC-Berkeley. His interest in coal mining stems from days as a chemical engineering undergraduate at BYU when he studied coal combustion. The explosion at the Sago Mine in West Virginia that killed 12 was the final impetus for this study.
Madsen’s data did not allow him to determine definitively why or how mining companies learn from error. But he offered his opinion as to why.
“We recognize coal mining as a relatively dangerous industry, but if you look at the numbers, any given mine is going to experience a fatal accident only once every 250 years, and it can be easy to forget exactly how dangerous it can be,” he said. “So when miners die, it serves as a wake-up call to others, helping them remember that this IS a dangerous industry, and they need to pay extra attention to the safety procedures they already have in place.”
As for why those memories fade over time, Madsen pointed out that annual employee turnover in the mining industry is 37 percent. When employees leave a mine that experiences a fatal accident, their experience and memory of it leaves with them.
To improve organizational learning around disasters, Madsen’s study prompts him to make two recommendations:
- Organizations should seek to learn more from minor accidents to prevent larger ones. One year after a minor accident, the learning it prompted had dissipated, on average.
- Post-accident investigations should focus more on determining exactly what happened and less on punishing those perceived to be at fault.
“I hope that accident investigations will become a little less punitive and a little more open to learning and recognizing that the causes of these accidents are many, and any one individual person who might have made a mistake wasn’t solely responsible for these accidents,” Madsen said. “If accident investigations take that approach then we’ll be able to understand these domains better and actually improve safety even more as a result of these accidents.”
Madsen’s future research will delve into mine disasters in more detail and is also expanding to look at airline disasters.
