A workers compensation claim is a lag indicator. By the time you are counting them, you are measuring failure that has already occurred, on a delay of months, sometimes years.
Last week's argument was that prevention has a reporting problem rather than an evidence problem, and that the way out is to measure exposure rather than harm. This week, what that actually looks like.
Week four of National Safe Work Month is about action across all workplaces, which is fitting, because none of what follows requires a budget, a platform, or a consultant. Most of it is data your organisation already holds and does not look at.
Lag, and lead
A lag indicator tells you what happened. Claims, turnover, absence, incidents, complaints. All useful, all retrospective, and all describing a situation you can no longer influence.
A lead indicator tells you about conditions now, in the hope of changing what happens next. For psychosocial risk, most lead indicators fall into three groups.
One: what the work is doing
These are measures of exposure, taken from operational data rather than from surveys.
Overtime, and where it concentrates. Not the total, the distribution. A team average of four hours can conceal two people carrying twenty.
Vacancies open longer than a quarter, and who is absorbing that work in the meantime.
Span of control. How many direct reports does a manager have, and at what point does noticing anything become impossible?
Leave balances that keep growing, especially among people who are performing well.
Rostering patterns, consecutive shifts, and how much notice people get of changes.
Exposure to aggression from customers or the public, which in many sectors is recorded somewhere and reviewed nowhere.
None of these require anyone to disclose anything. That matters, because the measures that depend on people telling you they are struggling will always undercount, and will undercount most in exactly the environments where the risk is highest.
Two: what you are doing about it
Process measures, which are unglamorous and carry more weight with a regulator or an insurer than anything else on this list.
Proportion of teams with a psychosocial risk assessment completed in the last twelve months.
Time from an issue being raised to something about the work changing. This one is uncomfortable and it is the single most revealing number here.
Proportion of leaders who have had practical training, and whether anything reinforced it afterwards.
Whether identified risks have named owners and dates, or whether they sit on a register nobody revisits.
Three: what people report about the conditions
Survey data, but asking about the work rather than about the person.
The difference matters. "How are you coping?" produces a wellbeing score, which tells you about individuals and invites the conclusion that some of them need more resilience. "Do you have enough information to do your job well?" produces a measure of role clarity, which tells you about the work and points at something you can change.
The same applies to questions about workload, about how much say people have over how work is done, about whether support is available when things go wrong, and about whether people believe raising a problem is safe. All of those map to named psychosocial hazards. None of them ask anyone to describe their mental health.
The trap worth naming
There is an obvious way to get this wrong, and it is worth saying out loud before anyone builds a dashboard.
Monitoring conditions is not the same as monitoring people. Keystroke logging, activity tracking, sentiment analysis of internal messages and similar tools are sometimes sold as psychosocial risk management. They are not. Intrusive surveillance is itself a named psychosocial hazard, which means a poorly designed measurement programme can create the exposure it was bought to detect.
The test is straightforward. Does this measure the work, or does it measure the worker? Overtime distribution measures the work. Reading someone's messages measures the worker.
One more caution. Nothing on this page is a screening tool and none of it identifies individuals who are unwell. These are workplace patterns, they belong in aggregate, and the moment anyone starts using them to form views about a particular person, they have stopped being a safety measure.
Worth asking
1. Of the measures above, how many does our organisation already collect somewhere? The answer is usually more than people expect.
2. Who looks at them together? Payroll holds overtime, HR holds vacancies, and nobody reads them as a risk picture.
3. How long does it take us, from an issue being raised to something changing? Would we be comfortable reporting that number?
4. Does anything we currently measure tell us about conditions before somebody is harmed? If not, we are reporting history.
The point
By the time a claim is lodged, the harm has happened, the person has carried it for months, and the organisation is paying for something it could have seen.
The patterns that precede it are visible well in advance, in data most organisations already hold. What is usually missing is not the data or the budget. It is someone whose job it is to look at it as a picture of exposure, and the authority to change the work when it says something.
Stay Curious,
Genevieve
Knowing what to look at is the easy part. Oasis Mind Body® works with organisations on the harder one, which is building leaders who can act on what they see.
This article is general information about work design and leadership capability. It is not legal advice, work health and safety advice, or clinical advice.
