No doubt the data will, if properly analyzed, allow OSHA and others to discern trends that could help head off future injuries and deaths. Perhaps artificial intelligence can analyze the data more quickly. But there remain differences between the data points to be collected—employee job title, date of injury or illness, location and type of injury and a very brief recount of what happened—and circumstances that can reveal the full story of what and why accidents occur.

Like any journalist researching stories, on the OSHA website, I look at what a company’s records show as a way of knowing its history of interactions with OSHA and whether there were prior injuries, fatalities or penalties that reflect on the employer or reveal a pattern.

The ever-more-robust OSHA data sets are invaluable. The agency’s fatality records, which in recent years have included names of those who died, are especially valuable in helping find brief descriptions of incidents that tell you more about what happened. You can triangulate the OSHA information with website news reports, GoFundMe pages and police accounts to fill in some of what occurred and how it affected the employee’s work and family life.

In some cases, simply repeating an employer’s inspection and violations data adds important context. I also especially appreciate published decisions of the OSHA Review Commission, the quasi-judicial entity that hears appeals in 23 states where the U.S. Labor Dept. enforces workplace safety (ENR 8/7-14/23, p.16). Commission judges evaluate testimony and evidence and, as I recently wrote, their decisions are often the best and most easily obtained detailed public discussion of a safety breakdown.

But data knows no loyalty. It is available to journalists or critics or rivals of an employer who may use it to create the impression that an injury or fatality was the inevitable culmination of a sloppy or callous company’s lack of concern for its staff safety. That’s often how it’s depicted in news reporting. One attorney even advises employers that had been in the habit of being scrupulously honest about details in their OSHA 300 forms to be more careful now about oversharing information.

The obverse is another favorite narrative, preferred by employers, of how employees somehow failed to adhere to safety rules—victimizing themselves through carelessness.

What advances the cause of safety better than data points on a spreadsheet, often taken out of context, is the deep story that can be at least partly revealed by an OSHA investigation. That story, not always sought by the agency, goes beyond tracing back to the root cause of the mishap and involves attitudes and conditions in the workplace.

The problem is that obtaining the full version of all results of an OSHA investigation, such as interviews with employees soon after the accident and photos of the scene that are beyond the brief summaries published—requires a Freedom of Information Act request. These often cannot be fulfilled for long months, or even years, as OSHA negotiates with employers over the final penalty and agency attorneys redact names.

Much prodding is often needed to obtain these records. Why not just redact names automatically and post the OSHA reports as a matter of practice, even years after an accident? What those reports contain—an investigator’s interview notes and onsite photos—could reveal deeper truths behind accidents. Publishing more data can help, but without deeper investigation, data points can be used to tell stories that aren’t true.