Suggesting that the agency pays the PAYE from April 2026 when Joint & Several Liability takes effect has created a storm in the industry it seems, so, with so many reactions -and plenty of inaccuracies -circulating about our proposals, I want to do two things: firstly, provide some context, and secondly, provide the transparency others claim to offer but fail to do so.
There are two fundamental principles that we’ve seen in play during many of the changes faced by the sector.
The first of these, is risk.
Risk is the major factor that drives decisions through the supply chain. These decisions are often made from the top of the supply chain down as any fallout ultimately hits the biggest brands in that chain which are often the large companies sitting at the very top.
The first example goes back to the introduction of the Agency Worker Regulations in 2010. At that point, two clear options were on the table: either offer workers comparable pay, or rely on the so-called Swedish Derogation - which, in simple terms, removed the obligation for comparable pay.
From day one, Professional Passport argued that comparable pay was the only sustainable option. Swedish Derogation, in our view, carried significant reputational risks for end clients and would ultimately be rejected. This position, though contrary to many others in the market - including some of the very voices now making accusations - was seen as disruptive and unhelpful at the time.
Events quickly proved us right. Almost immediately after the regulations came into force, headlines emerged accusing brands that had chosen Swedish Derogation of failing their workers. Only two or three brands faced this backlash, but it was enough to make the risks clear. Those providers who had insisted Swedish Derogation was the only viable route were forced into a rapid pivot, and the model collapsed so completely that it was later removed as an option, quietly and with little notice.
Our position may have been unpopular, but it was driven by a simple truth: clients don’t like risk. Whether the risk is financial, reputational, or purely theoretical, they will go out of their way to avoid it.
A more recent example is Off-Payroll Working.
At that time, many providers had substantial accountancy divisions, often far larger than their umbrella operations.
Those involved, and affected by these changes, will all tell how almost overnight the market shrank by over 50% as clients could see a risk. Many didn’t understand this risk, which further fuelled the decisions.
The beneficiary of this was the umbrella market as the clients wanted PSCs out of the supply chain. Umbrellas were seen as a safe alternative as they removed these risks.
Interestingly the market for PSCs has never really recovered even though it has been proven that the risks now are both negligible and manageable.
Now let’s address he second factor that drives decisions through the supply chain - Control.
Where there is risk and no control over that risk the supply chain will simply decide to remove it, a lesson reinforced by Off-Payroll Working.
Legal professionals would never suggest to clients that they take on risks that they have no control over, where an alternative exists.
Once these decisions are made, they are not easily overturned and so the real challenge is how to stop them being made in the first place.
If you read some of the reactions to our suggestion they centre on the umbrella losing control if they don’t pay the PAYE themselves. The reality is that parties further up the supply chain face the same issue: they now carry a risk they cannot control. Without a workable compromise, the likely market reaction will be a shift toward agency PAYE, a move that neither umbrellas nor agencies want.
We have also been accused of working against the sector by putting this proposal forward. There have been many occasions where people predict the death of the umbrella only for the market to grow and expand. Change is always painful but sometimes necessary.
I'm sure many will remember the fear when the expense rules changed. Headlines like 'What's the point of umbrellas now?' and 'It's the end of the umbrella' were regular features across industry publications. Yet despite these predictions, the umbrella market has continued to grow.
Addressing the harsh reality of JSL is a pivotal moment for the market as if we lose market share it will take years to recover. For umbrellas to thrive we need a volume market, and without this it becomes increasingly difficult to make the model work.
The only viable path forward, I suggest, is to prove we can eliminate risk while relinquishing some of that control.
I understand why many praised JSL as a great outcome and claimed victory in securing this change but I would argue that the real victory was actually retaining the Umbrellas PAYE reference for reporting earnings. The alternative of using the Agency PAYE reference would have been a complexity that is best avoided. Yet this success comes at a price, one that many, including some of those who advocated for these changes, had not fully appreciated.
Now let’s look at the current market offerings:
Accreditations, in simple terms offer no protection against JSL, but still have a role to play in the overall jigsaw. Whilst JSL is one aspect of the legislation, there are many others, such as The Criminal Finance Act, that require a much broader due diligence process, significantly more than a payslip check or RTI Verification. So, we can’t allow ourselves to become fixated on one single aspect and lose focus of the bigger picture. While detailed analysis of an umbrella’s operational processes and procedures still has an important role to play, this is not relevant in the context of JSL.
Payslip checks. Whilst these do offer some insight it is vital this is not overstated. Many of the payslips we have seen can be best described as false, even though the calculations presented would pass a payslip check. The first piece of critical data in the variable equation is what was sent to the provider, not a test against reported gross taxable earnings stated. Validating that the numbers align to what a worker receives is also a critical component. Without these 2 fundamental pieces of information the system can be duped, resulting in false hope. This can also have a major impact on the stated liabilities through RTI, which brings me to the next point - RTI Verification.
If I remove all aspects of where this could be manipulated and just look at the reality of what this means under JSL with a simple worked example it fails to remove risk.
My main issue with this is the lack of transparency about what this really means. The marketing of these checks claims to solve the issue and we have seen words like ‘guarantee’ being used. This is misleading and this lack of transparency brings significant risk and when discovered could trigger the same fear of liabilities highlighted above, potentially leading to the wrong decision being made.
Don’t get me wrong; this can have a role where companies are happy with the underlying risk but follow your own stated objectives of being open and transparent.
At the risk of teaching granny to suck eggs, although from some comments, including those by so called experts, it does appear this is not fully understood.
So, let me explain in an example where I have rounded numbers for ease but the key point is the principle in play.
Let’s use an agency with workers operating through an umbrella company that results in a weekly PAYE bill of £20,000.
PAYE liabilities accumulate over a period of weeks that make up monthly cycles. The settlement of each monthly cycle is due on the 22nd of the following month.
When PAYE is paid it generally takes 7-10 days for this to show on the providers HMRC portal, but for the purposes of this example I will ignore that time lag.
So, on a weekly payroll multiple pay runs have been reported and RTI returns submitted to support these.
The day before payment is due there would be an outstanding liability in this example of around £160,000.
Once payment had been made there is still an outstanding amount due, as pay runs have already been reported for the second period that don’t need to be settled until the 22nd of the next month. So, the day after there still could be around £60,000 of liability reported but not yet due and not paid.
So, the reality for this agency is that where they are happy to carry a rolling potential liability of between £60,000 - £160,000 this works fine.
Now this is where I refer to collective liability, and to be clear this is my phrase and in no way relates to the application of the legislation. It could also be referred to as a rolling liability; what it is called is unimportant because the important point is there is a potential liability.
If anything happens that causes that umbrella to fail, that is nothing to do with this agency, there is now an unpaid liability. The agency has sent all the money to the umbrella but this hadn’t been paid as it’s not due. So, if this event happens, the agency will face an additional bill of between £60,000 and £160,000 depending on timings of the event.
This is an additional cost as the PAYE monies have already been sent with the original funds.
Of course, if the umbrella pays the PAYE immediately then this significantly reduces any risks although there will be some, due to the time lag of systems to record this payment.
With a monthly verification you will have one of 2 potential outcomes;
Confirmation that PAYE has been paid for the first period and your rolling liability has reduced.
The one no one wants; the PAYE has not been paid and so expect a bill from HMRC for the £160,000.
Checking after the fact fails to address the reality of the new world with JSL.
Our proposal of the agency making the payment each week satisfies the risk and control aspect from their perspective, but I accept moves this risk back to the umbrella which is what seems to be the issue.
There is no simple answer to removing this from the umbrella but it is manageable.
The umbrella can view any credits that have been applied through their portal, and where workers are paid on the Friday the agency could be required to make payment on the Monday. The umbrella should be able to verify this payment within 7-10 days through their portal.
So yes, there is a liability being held but that liability is significantly less than the liabilities where credit terms are offered, and many umbrellas are willing to take these risks on, often without supporting insurance cover.
These risks are also insurable risks if the agency goes into administration.
The umbrella retains responsibility for all the work they currently do and those who present this as a threat to the umbrella are simply scaremongers - my view is the absolute opposite.
By removing the risk and being prepared to relinquish some control you are placing the umbrella sector in the strongest possible position to continue to grow and flourish. Failing to be transparent and masking risks is what will result in a diminished umbrella market.
I find it remarkable that representatives of umbrellas are suggesting that the agency is the issue, essentially attacking the very companies that their clients serve., I fail to see how this position in any way helps or promotes an umbrella company’s business.
Some are suggesting that we are only offering this to sell a new service, yes we have a service to deliver this but there seems to be a presumption around costs - these assumptions are simply wrong and seem to be fuelled by parties with vested interests.
And to be absolutely clear our new service can offer payslip checks and verifications but we would only do this with complete transparency of what that really means, as explained above.
We decided to promote this offering as our primary solution to JSL as it is, in our opinion, the only approach that fully removes risk and addresses control issues.
So, the simple fact is there is now a choice between either verifications that do hold risk or agency-paid PAYE which removes the risk.
