Why Regulating Umbrella Companies Would Have Been Better Than Joint & Several Liability

The UK contractor industry is once again being reshaped — this time by Joint & Several Liability (JSL) in the Finance Bill.
According to HMRC, JSL is designed to tackle tax non-compliance in labour supply chains by allowing unpaid tax liabilities to be passed up the chain to agencies and end-clients.
But while JSL may help HMRC recover lost revenue, it raises a serious question:
Why introduce yet another complex liability mechanism instead of properly regulating umbrella companies in the first place?
A long-standing problem, repeatedly side-stepped
Umbrella companies have been central to the contractor labour market for decades, particularly since IR35 reform pushed large numbers of contractors into PAYE employment models.
Yet despite their prevalence:
Umbrella companies are not defined in statute
They are not licensed
They are not directly regulated as a sector
Instead, umbrellas sit in a regulatory grey area — treated as employers for tax purposes but largely invisible as a distinct labour market intermediary.
This lack of definition has allowed non-compliant operators to flourish, hiding behind labels such as “payroll provider” or “employment facilitator” while engaging in skimming, unlawful deductions, or disguised remuneration.
Joint & Several Liability: addressing symptoms, not causes
Joint & Several Liability does not stop non-compliant umbrellas from forming or operating.
What it does instead is:
Shift financial risk away from the umbrella
Place it onto recruitment agencies and end-clients
Force supply chain participants to carry out defensive due diligence
In effect, JSL assumes that non-compliance will continue — and simply decides who should pay when it does.
This is a reactive, not preventative, approach.
Bad umbrellas can still:
Enter the market
Recruit workers
Operate until detected
Collapse and re-emerge under a new name
JSL may improve tax recovery, but it does little to clean up the underlying market.
Regulation would have been simpler — and more effective
By contrast, statutory regulation and a clear legal definition of umbrella companies would attack the problem at source.
A regulated umbrella sector could include:
A clear definition capturing all umbrella-like models
Mandatory registration or licensing
Minimum compliance standards
Transparent payslip requirements
Direct enforcement powers against offending umbrellas
This would prevent non-compliance before workers are harmed and before tax is lost.
It would also reduce the need for increasingly complex mechanisms like JSL, which add administrative burden without eliminating bad actors.
Even the government now accepts this logic
The government itself has acknowledged the regulatory gap.
Umbrella companies are due to fall under the remit of the Fair Work Agency, expected to begin operating in spring 2026, with full regulation likely from 2027 following consultation.
This raises an uncomfortable question:
If regulation is the right answer now, why wasn’t it pursued earlier — before introducing Joint & Several Liability?
Complexity is not a substitute for control
The contractor market is already burdened by:
IR35 reforms
Off-payroll working rules
Supply chain due diligence
Status determinations
Now, Joint & Several Liability
Adding further complexity without fixing the root regulatory gap risks making the system harder to navigate for compliant businesses — while determined bad actors continue to adapt.
Proper regulation would not eliminate every problem overnight, but it would:
Simplify enforcement
Clarify responsibility
Protect contractors directly
Reduce the need for ever more elaborate tax measures
Conclusion: a missed opportunity
Joint & Several Liability may help HMRC recover tax, but it does not fundamentally fix the umbrella market.
A statutory definition and regulatory regime would have been:
Cleaner
Fairer
More preventative
Less disruptive to compliant supply chains
Instead of continually layering new rules on top of an unregulated sector, the government could have — and arguably should have — tackled the problem at its source.

