Wellbeing measurement

There is no prescribed instrument, no required frequency and no official threshold to hit. What there is, in the Management of Health and Safety at Work Regulations 1999, is a duty to make a suitable and sufficient assessment of the risks to employees' health, and HSE applies that duty to stress. Regulation 3(3) requires review when there is reason to suspect the assessment is no longer valid, or after a significant change. That is a trigger, not a date.

So the useful test of a measurement approach is not how sophisticated it is. It is whether it leaves something dated, specific and reviewable. Annual survey or continuous evidence works through the four questions any record has to be able to answer, and which approach is better at which.

This surprises people, and it comes from the regulator rather than from us. HSE says of its own instrument that a survey is not an essential step and for smaller organisations it would not be proportionate to run such a survey, particularly where the same data can be gathered in other ways.

What it publishes instead is a Talking Toolkit: six conversation templates, one for each of the six Management Standards, built on the basis that structured conversations can gather what a larger organisation would gather through a survey. Recorded, dated conversations are evidence.

HSE also says it does not expect an employer to meet all six standards, and no current HSE page sets a target percentage. The 85% and 65% figures that circulate as Management Standards targets came from a 2004 pilot whose own authors described the thresholds as based on research that was indicative rather than grounded. That is why nothing on this site gives you a score.

Four of them. Each does one thing, states its limits, and sends nothing anywhere.

• Survey gap clock. Enter the date of your last survey and the cadence you meant to keep. It counts the gap and nothing else: no score, no dataset, no prediction.
• Psychosocial risk checker. The six HSE Management Standards, two questions each, and what evidence in each area usually looks like. Produces no score, because HSE sets no threshold.
• UK sickness absence rates. The 2025 ONS release, eleven breakdowns, every figure read out of the published workbook. National context, and the page is explicit that it is not an employer benchmark.
• Mental health map of England. Four ONS4 personal wellbeing indicators across 150 English local authorities. Population context for the area people live in, which is not a description of anyone who works for you.

The two that use public data, the absence explorer and the map, carry the same warning in different words: population data describes a population. Your employees' postcode is context, not a diagnosis, and a national industry rate is not a benchmark for your organisation.

Both public-data tools are reproduced from published releases without recalculation, and both are built by a script that reads the source rather than by anyone transcribing figures.

Sickness absence. Office for National Statistics, Sickness absence in the UK labour market, 2025 edition, released 1 May 2026, from the Labour Force Survey, under the Open Government Licence. A working day is 7 hours 30 minutes. ONS labels these official statistics in development, there is a reweighting discontinuity between 2018 and 2019, and industry, occupation, sector and workforce size are self-defined by the person answering rather than taken from an employer record. Cells based on between 3 and 25 responses are flagged on the page itself.

The map. ONS4 personal wellbeing indicators (anxiety, life satisfaction, happiness, worthwhile), 2022/23, at upper tier local authority level, from the OHID Fingertips extract, 150 areas with published data. Two things about it are worth knowing before quoting it. The confidence intervals are wide: 138 of the 150 areas have an interval containing the England value, so the map shows a spread rather than a ranking. And the geography vintage used for the extract has since been retired, so the published figures cannot be reproduced from today's Fingertips API; re-running the same indicator on the current Counties and UAs geography returns 149 areas rather than 150, with the range and the ratio unchanged.

The absence data can be re-derived from source on demand, and the page fails a check if it has drifted from the workbook. The map is a fixed extract, so it carries its vintage on the page instead.

It cannot tell you about an individual. Nothing here produces an individual result, and the check-in below never shows one.

It cannot tell you that you are compliant. Whether an assessment is suitable and sufficient is a judgement about your particular workplace, and no tool has enough information about yours to make it.

It cannot predict. Not resignations, not absence, not mental ill health, not claims. Where a number would have to be invented to make a prediction work, we have left the number out, and the individual pages say where and why.

And a measurement nobody acted on is not evidence of care. It is evidence that you looked, which is the smaller half of the question.