Safety and OSHA News

State study finds 2.5 times the number of worker injuries as federal count

Researchers have once again shown that federal counts of workplace injuries don’t show the whole picture – not by a long shot. 

Michigan State University researchers found that over a three-year period, workers in the state suffered two-and-a-half times the number of crushing injuries than reported by the U.S. Department of Labor’s Bureau of Labor Statistics (BLS).

For 2013 through 2015, BLS reported 1,260 crushing injuries in Michigan.

Using records from all 134 of Michigan’s hospitals, the Michigan Workers’ Compensation Agency and Michigan’s Fatality Assessment Control and Evaluation (MIFACE) program, the MSU researchers counted 3,137 work-related crushing injuries, including two fatalities.

The researchers’ finding that BLS undercounted crushing injuries by 59.8% is similar to what they’ve found previously looking at other types of workplace injuries: amputations were undercounted by 59%, burns by 69% and skull fractures by 54%.

Another difference between the two counts: BLS found the rate of crushing injuries had declined during the period, while the researchers said the rate actually went up.

What are some reasons for the BLS undercount?

  • BLS counts only injuries which caused one or more days away from work
  • employers aren’t providing complete reporting
  • employees may not report for fear of retaliation or because of incentive programs rewarding lack of injuries, and
  • BLS excluded self-employed, household employees and workers on farms with less than 11 employees.

The researchers say not only does their count provide a more realistic picture about workplace injuries, it also allows Michigan OSHA (MiOSHA) to better target workplaces. Reporting by the researchers led to 77 MiOSHA inspections. At 79% of those companies inspected, the hazardous condition that caused the crushing injury hadn’t been corrected, even though the inspections occurred three to six months after the injury. Companies inspected faced MiOSHA fines up to $37,000, in many cases including a citation for not reporting the injury.

Is the Michigan count complete? Maybe not, said the researchers, because there are still limitations to their system. Sometimes the medical record information was limited. If an injured Michigan worker was treated in a hospital in a nearby state or in an urgent care center or physician’s office, it wasn’t counted.

Could the federal government conduct a similar workplace injury count? Good question. A National Institute for Occupational Safety and Health (NIOSH) program, the National Electronic Injury Surveillance System, reviews records from a sampling of hospital emergency rooms across the nation. Since it’s a sample, it’s not a complete count like the Michigan study.

What’s the ultimate goal of the Michigan count? The researchers say it’s to recognize and prevent work-related injuries.

Here’s a reason why all employers should care whether workplace injuries go unreported. As the Michigan study shows, not reported also means not inspected. And in 79% of those cases, hazardous conditions weren’t corrected even after a worker suffered a crushing injury.

It’s confirmation that some employers choose to continue to roll the dice regarding worker injuries even after serious ones that require a hospital visit. The reason they do so is to keep costs down. Honest employers that spend the money to eliminate injuries in the first place and accurately report when they do happen are put at an economic disadvantage compared to the companies that are taking their chances with safety.

Should more states set up programs like the one in Michigan to accurately count workplace injuries and, in the process, find employers who are hiding them? Let us know what you think about that and anything else related to the Michigan study in the comments.

Print Friendly

Subscribe Today

Get the latest and greatest safety news and insights delivered to your inbox.


  1. Tom Rezner, Ph.D.,SPHR, CSM says:

    There are specific regulations about reporting. One of the big problems is the feds and states don’t see things the same way. in terms of reporting what has to be reported and when. The federal reporting tries to catch many non worked related cases which either are paid by employers to prevent hassle and maintain employee relations or contested for fraud. Cases a state Judge rules are not work related or a reputable doctor rules not work related or not lost time the feds still want to claim the opposite. Not every company has highly skilled Safety and HR people who know the Regs and report accordingly. One thing for sure Compliance is taught at college.

  2. I think they hit good reasons why the reporting doesn’t isn’t done. At my level (the site level) I think it has a lot to do with punishment for Recordable injuries.
    I was at a site a few years back where we went 4 years and over a million safe man-hours without a recordable injury. We had one where an employee took a short cut, which he knew to be wrong, and injured himself and had to have sutures to close a wound. After the review I was to be replaced as Safety Manager. The reason that the VP told me, to my face, was “What do you expect? You had a Recordable.”. I felt I was punished for not “turning” this in to a non-recordable.
    I’ve heard similar stories from a lot of other site safety managers over the years. I’ve had other Safety Managers tell me that there is no injury they can’t make non-recordable.

    • Tom Rezner, Ph.D.,SPHR, CSM says:

      There are always idiots in Mgt who don’t realize that any run of luck has to come to an end. You don’t replace someone for 1 accident unless they created the situation. You replace them when a trend develops and they can’t do something about it.

  3. David Combs, MS, CSP, CFPS, CPSI says:

    Just had this discussion last week at the OTI and many in the class said that I was incorrect. Every December I get a lot of request for loss runs so clients can fill out their 300 log. I always tell them that if their using their WC loss Hx for their OSHA logs, they are over reporting injuries. The criteria for a WC claim is much lower than an OSHA recordable. If an employee gets injured on Friday but is back to work on Monday, there is no recordable (depending on the severity) however, there will likely be a WC claim to help cover the costs of immediate and future treatment of the injury. Also, many of my client’s employees suffer injuries while in a travel status (car accident, slips in the hotel shower etc). Again, these are not OSHA recordable but very well can be WC claims.

  4. Philip Hannifin says:

    Each administration has under reported for years. If they were reporting properly, it would continue to show disastrous safety results. OSHA compliance does not equal safety. Here is one regulatory agency that needs to show improvements to justify the lifelong middle administration jobs where they do nothing proactive and live off the hard working tax payer and businesses they threaten with punishment that only really comes after a catastrophic event. If they don not show improvements, they cant justify their existence.

    • Tom Rezner, Ph.D.,SPHR, CSM says:

      The numbers don’t lie but they don’t tell the whole story. Workers Comp pays for many incidents that probably aren’t worked related accidents. Many minor skinned knuckles, floor burns, scratches, sprains, bruises get paid for cash or as a deductible, but smart employers report them to Workers Comp as information only so if they go bad there is a record. Many employers send minor first aid cases off to the doc and as a result there or Doctor bills and a Workers Comp file is created. In general I would think the over cautious send to the Doc and the sensible in house/no record care of injuries is probably a wash. When one looks at the OSHA Log OSHA’s Regs are designed to create accidents on paper that are either not work related or are not as severe as they really are. The obvious reason for this is Safety Professionals are doing to good a job and OSHA needs accidents to justify employee jobs and OSHA’s existence. The one set of number I have some faith in is Workers Comp PAID lost time. Work Comp has a waiting period before you can get paid.

    • I think it is less about under reporting and more about a fatally flawed study. These researchers are comparing apples to oranges. The BLS data contains only OSHA recordable cases with lost time. Many of the cases the researchers compared that number to were cases where employees sought medical attention but presumably received first aid, diagnostics, or for some other reason the injury incurred no lost time. Some of them were OSHA recordable, others not. That does not mean they were “under reported” it just means they are counted in a different bucket. Interestingly I reviewed my OSHA injury data from last year and the total number of injury reports was almost 3 times the total number of recordable injuries which was about 1.75 times the number of lost time claims. So what the Michigan research shows is that there are generally more total claims than there are claims with lost time. Wow, big Eureka moment there.

  5. Ah the dangers of comparing Safety data to Workers’ Compensation data. Clearly the researchers at Michigan chose not to consult professionals from either field to review their study. Or perhaps they did but chose to ignore their input because it did not support whatever the end goal of the study was. Who knows.

  6. Fred Hosier Fred Hosier says:

    Editor’s note: While the research does not calculate whether under-reporting or the difference between OSHA recordable and workers’ comp compensable injuries has more of an impact, consider this part of the research: “Reporting by the researchers led to 77 MiOSHA inspections. At 79% of those companies inspected, the hazardous condition that caused the crushing injury hadn’t been corrected, even though the inspections occurred three to six months after the injury. Companies inspected faced MiOSHA fines up to $37,000, in many cases including a citation for not reporting the injury.” That’s 77 recordables that weren’t recorded for just one specific type of injury: crushing. Their previous research on other types of injuries had similar results. That statistic gives some idea about the impact of under-reporting.


  1. […] years. While their study is limited to Michigan, the issues are similar in Alabama. What does the data collected in Michigan […]

Speak Your Mind