Repeat enforcement: what four years of HSE notices reveal about the limits of point-in-time supplier safety checks
Most supplier safety assurance in UK construction rests on a single assumption: that a company which passed an assessment is a company that is safe to engage. Accreditation schemes, pre-qualification questionnaires and approved-supplier lists all work this way. A contractor submits evidence, an assessor reviews it, a certificate is issued, and the contractor joins the list. The assessment is a photograph. The engagement that follows is a film.
We wanted to know how much that distinction matters in practice, so we analysed every Health and Safety Executive enforcement notice issued to a UK construction company over a four-year period and matched it back to the Companies House register. The pattern that emerged is not that enforcement is common. It is that where enforcement happens, it frequently happens more than once, to companies that remain on the register and available for hire.
What we looked at
We examined 4,457 HSE enforcement notices issued between 1 January 2021 and 31 December 2024 where the HSE classified the recipient's industry as construction. Of those notices, 2,095 were immediate prohibition notices, the instrument the HSE uses when an activity carries a risk of serious personal injury and must stop straight away. The remainder were predominantly improvement notices, which require a specified contravention to be remedied within a set period.
Those notices were issued to 2,491 distinct companies. We matched each notice to the Companies House register by company name, retaining exact-name matches, matches confirmed by registered postcode, and matches to a company's recorded previous names. We then looked at each company's current status on the register.
We have deliberately not named any company in this analysis. Individual enforcement notices are already published by the HSE and are a matter of public record; our interest is in the aggregate pattern, not in any particular contractor.
The headline finding
Of the 2,491 companies that received an enforcement notice during the period, 2,088 remain active on the Companies House register today. A further 400 are in liquidation, in administration, dissolved, or subject to a proposal to strike off.
Among the 2,088 still-active companies, 1,247 had received at least one immediate prohibition notice. In other words, roughly six in ten currently active construction companies with an enforcement history have at some point been ordered by an inspector to stop work immediately because of a risk of serious injury.
Then we looked at repetition, and this is where the assessment-as-photograph problem becomes concrete.
Separating repeat enforcement from a single bad inspection
A company can receive several notices on the same day. An inspector who attends a site and finds four distinct contraventions may issue four notices, all dated identically. Counting those as four instances of enforcement would materially overstate how often companies come to the HSE's attention, so we separated the two cases.
Of the 2,088 active companies, 823 received more than one notice during the period. But 386 of those received all of their notices on a single date, which is one inspection rather than a pattern. Excluding them leaves 437 active construction companies that were subject to HSE enforcement on more than one separate occasion between 2021 and 2024.
Broken down by the number of separate occasions on which enforcement occurred:
| Occasions | Companies | Share |
|---|---|---|
| One | 1,651 | 79.1% |
| Two | 344 | 16.5% |
| Three or four | 81 | 3.9% |
| Five or more | 12 | 0.6% |
Of the 437 companies with enforcement on more than one occasion, 151 had incidents at least 90 days apart, and 83 had incidents separated by a year or more.
That last figure is the one worth sitting with. Eighty-three active UK construction companies came to the attention of HSE inspectors, had enforcement action taken against them, and then came to the attention of inspectors again more than twelve months later. A full annual assessment cycle passed between the two events.
Where the notices were issued
The HSE assigns each notice to a region. Across the four-year period, for construction companies:
| Region | Notices | Prohibition notices | Companies | Still active |
|---|---|---|---|---|
| London | 1,628 | 842 | 636 | 536 |
| East & South East | 1,242 | 654 | 595 | 493 |
| Yorkshire & North East | 993 | 543 | 366 | 318 |
| North West | 904 | 422 | 317 | 256 |
| Midlands | 807 | 442 | 304 | 254 |
| Wales & South West | 770 | 344 | 327 | 277 |
These are counts, not rates. We have deliberately not expressed them as a proportion of companies in each region, because the number of construction companies registered to an address in a region is a poor proxy for the amount of construction work carried out there. A contractor registered in one region routinely works in another. Read the table as a measure of where enforcement activity occurred, not as a league table of regional safety performance.
What this does and does not tell you
An enforcement notice is not proof that a company is unsafe to engage. Construction is a heavily regulated industry with genuine hazards, and inspection is concentrated where the risk is greatest. Larger contractors running more sites will, all else equal, be inspected more often than a two-person firm. A single improvement notice, promptly complied with, may say very little about a contractor's overall safety culture, and some of the best-run firms in the industry appear in these figures.
There are limits to the data itself, which we would rather state than have inferred.
Our HSE coverage begins in January 2021, so a company with a poor record before that date and a clean one since will look identical here to a company with no history at all. We have excluded Scotland entirely, because our HSE coverage for Scotland only begins in December 2025 and including a partial series alongside four complete ones would produce a misleading comparison. We have also stopped the analysis at the end of 2024 rather than running it to the present, because enforcement data for recent periods is still settling and we would rather report four complete years than five uneven ones. Every one of these decisions makes our figures smaller than the true position, not larger. Treat the counts above as a floor.
Finally, our matching is by company name rather than company number, because the HSE publishes the name of the notice recipient rather than a registered number. We retained only high-confidence matches, but name-based matching cannot be perfect, and companies with very similar names present a genuine difficulty.
Why a point-in-time check cannot see this
Return to the 83 companies with enforcement more than a year apart. Consider what a pre-qualification assessment would have shown at the midpoint. The first incident has been dealt with. The notice has been complied with. Whatever the assessor asks about is answered honestly, because at that moment there is nothing outstanding. The certificate issues. The contractor joins the approved list. Some months later, the second incident occurs.
Nothing has gone wrong with the assessment. The assessment did exactly what it was designed to do, which was to evaluate a company at a moment in time. The problem is that a moment is not the thing procurement teams actually need to know about. What matters is the trajectory, and a trajectory is only visible if you are looking more than once.
This is the case for continuous monitoring rather than periodic reassessment, and it is not an argument against accreditation. Pre-qualification establishes a floor. Monitoring tells you whether a supplier is still standing on it. The two do different jobs, and in most organisations only the first is being done.
The practical step
You can do a version of this analysis on your own supply chain without any software at all. Take your active construction and trades subcontractors, and search each one on the HSE's public enforcement register. Note not just whether anything appears, but whether it appears on more than one date. Then check the company's current status at Companies House, because a contractor in a proposal-to-strike-off state is a continuity problem as well as a safety one.
For a list of twenty subcontractors this is an afternoon's work. For two hundred it is not realistic to do once, let alone repeatedly, which is the practical reason most organisations check at onboarding and then stop.
Senserity monitors HSE enforcement notices and convictions, Environment Agency enforcement, and a further twenty-plus UK public and commercial sources continuously across a supplier portfolio, and flags changes as they are published rather than at the next review date. Our Operational and Compliance risk categories incorporate the enforcement history described in this article, and our procurement workflow is built around the distinction this analysis illustrates: the difference between what was true at assessment and what is true now.
If you would like us to run the checks described above across your own subcontractor list, we offer a one-time Supplier Risk Baseline Review covering up to 10 companies. Get in touch and we will show you what a portfolio-level view of your supply chain looks like.
Analysis based on 4,457 HSE enforcement notices issued to UK construction companies between 1 January 2021 and 31 December 2024, matched to the Companies House register. Company status as at July 2026. Scotland excluded. Source data: Health and Safety Executive public enforcement register; Companies House.