In labour-dominant services the public sector sets the pay, not the market. On a recompetition the workforce transfers on its existing terms, so an authority scoring pay properly is choosing to spend more money. Nothing published with PPN 026 says where that money comes from, and the safe assumption is that it is the authority's to find.
PPN 026 (Cabinet Office, August 2026) asks central government bodies to put a minimum 20% of the marks on social value for contracts of £5m and above, and 10% from £1m, for procurements commenced on or after 1 January 2027 (paragraphs 10 and 13). Most of the conversation since publication has been about the percentages and the narrowing of outcomes. This article is about a different number.
In services where most of the price is people (cleaning, catering, security, care, facilities, repairs), most contracts are recompetitions, and the workforce transfers to the winner under TUPE on its existing terms. So where the incumbent pays the statutory floor, an authority scoring the fair pay criterion (MAC 1c) properly is not asking the market to do something. It is deciding to pay more out of its own budget and inviting bidders to price it. A better workforce model costs six to twelve per cent on the bid, anchored on the Homecare Association's build-up of what direct employment actually costs, which on a £5m contract is £300,000 to £600,000 a year of the authority's money.
That decision gets taken, or ducked, before anyone drafts an award criterion. So the rule this article argues for is a short one: the budget decision comes before the MAC decision.
what a better workforce model adds to the bid price in labour-dominant services. On a £5m contract, that is £300,000 to £600,000 a year of the authority's money.
Anchored on the Homecare Association's direct employment cost build-up
On a like-for-like recompetition the people doing the work arrive on day one on the terms they already had, because that is what TUPE is for. The winning bidder inherits the payroll; it does not create one. Whatever the tender says about fair pay, the new contract starts from the old contract's wages, and moving off them costs money that has to appear in the price the authority pays.
Which means that in labour-dominant services the public sector largely sets the pay, not the market.
The budget decides what the workforce is worth, and the award criteria only distribute marks inside that envelope.
The evidence runs in both directions: the Homecare Association found 27% of homecare contracts in England priced below what direct employment costs at the legal minimum wage, which is buyers setting pay downwards, whatever their tender documents scored.
I watched this land about thirteen years ago. Two London boroughs were some way into a collaboration on parking enforcement, on the usual logic that one joint contract would cost less than two separate ones. It was abandoned when it became clear that one of the boroughs had already committed to the London Living Wage. Harmonising the workforce meant the other borough's costs going up rather than down. The saving that justified the whole collaboration turned out to be the gap between what the two authorities were prepared to pay their contracted staff.
of homecare contracts in England are priced below what direct employment costs at the legal minimum wage. That is buyers setting pay downwards, whatever their tender documents scored.
Homecare Association, Fee Rates for State-Funded Homecare
If a cleaner on a government contract moves from the statutory floor to the Real Living Wage, the return is about £1,500 a year in that person's pocket (Living Wage Foundation rates), and hours they can plan a life around. That is the policy working. It accrues to the worker rather than the authority, and PPN 026 says as much: the model exists to deliver good jobs, skills and opportunities in the communities where the money is spent. We say that plainly because the rest of this section is about money, and it is easy to slide from asking how to fund something into asking whether it deserves funding.
The commercial case for better pay is real but modest, and it lands on the supplier's side of the line. Cardiff Business School's survey of 1,532 Living Wage accredited employers found almost all reporting positive effects, mainly reputational, then in recruitment and retention, and the researchers' own summary is that the scale was often modest, a benign rather than a transformational change. Lower turnover reduces the contractor's cost base, not the authority's, which is why the argument rarely survives a budget meeting.
The authority's own case sits on the other side of the ledger. The Sector Pulse Check (Hft and Care England) found 43% of adult social care providers had closed parts of their organisation or handed back contracts because of cost pressures, with workforce costs the most cited pressure. And ADASS East's provider risk profiling already brings workforce turnover together with financial and safeguarding data to spot providers heading for failure. An authority weighting MAC 1c is buying against a failure mode its own colleagues already monitor.
of adult social care providers had closed parts of their organisation or handed back contracts because of cost pressures, with workforce costs the most cited pressure.
Sector Pulse Check, Hft and Care England
This is the question we cannot answer for anybody, and the one that decides whether PPN 026 does anything in labour-dominant services. £300,000 to £600,000 a year has to come from somewhere, and there are only so many places.
It can come from central funding, if central funding carries it. There is precedent for national money following a national pay commitment: Care Act statutory guidance already requires councils to set fee levels that let providers pay at least the minimum wage, and the adult social care Fair Pay Agreement is being funded in part through the local government settlement. Nothing of that kind has been published alongside PPN 026, and we would not plan on it appearing.
It can come from council tax, within the 3% core and 2% adult social care precept thresholds, which is a political decision rather than a commercial one.
It can come from the supplier's margin, which works once and not twice, and tends to end in a contract handed back.
It can come from scope, and this is the route we would watch most closely, because it is the easiest and the least visible: an authority can raise the hourly rate and commission fewer hours. The workforce is better paid and fewer people get the service. Nothing in the model would catch that, and every KPI would look fine.
Or it comes from nowhere, and the authority weights the MAC anyway. That produces one of three things: a good bid it rejects on price, an unfunded promise it never checks, or a supplier that promises the workforce the tender scored and delivers the one the price funds.
PPN 026 binds central government departments, executive agencies and non-departmental public bodies, and everybody else may adopt it (paragraph 4). "Everybody else" covers organisations with nothing much in common financially: NHS trusts, councils, schools, universities, police and fire services, housing associations. Their money arrives by different routes on different timetables, so there is not one adoption decision. There are thousands.
What they share is the incentive. An open day, a college partnership and a work experience programme cost a supplier very little, so weighting the access criteria (MAC 1a, 2a and 2c, which ask how a supplier will open the work up to people currently outside it) costs the adopting body very little. Weighting pay on a contract where the workforce transfers is not free at all. So the cheapest way to adopt PPN 026 is to adopt the parts that do not touch pay, and the result is a procurement that looks like PPN 026, reports like PPN 026, and leaves the pay of the contracted workforce where it was.
Whether that happens depends on money we cannot see from here. The 2026/27 to 2028/29 local government finance settlement distributes £83.5bn rising to £90.5bn (House of Commons Library), and the Local Government Association notes that 10.5% of social care councils still face a real-terms reduction in core spending power in 2026/27. The equivalent numbers for a department or an NHS trust sit inside its own budget rather than in a public settlement, so each body will do this sum for itself, with numbers nobody outside it can see.
The adult social care Fair Pay Agreement is the exception that shows the shape of an answer. There, pay is being settled nationally by statute and part-funded through the settlement. Everywhere else, PPN 026 asks a version of the same question and leaves it to whoever writes the budget.
Sub-criteria and practitioner guidance arrive in autumn 2026 (paragraph 12), and departmental mandates will follow. We would not wait for either to answer the funding question. Nothing published alongside PPN 026 points to money following the mandate, and our working assumption is that the guidance will say how to apply the criteria and stay silent on how to fund them. If that proves wrong and money arrives, nothing in this article is wasted. If it proves right, an authority that waited has a live tender and no answer.
So the challenge sits with each contracting authority now: cost the workforce the criteria describe, decide whether the authority can and will fund it, and take that decision with eyes wide open, everyone in the room including whoever owns the budget, before any MAC is selected. An authority that weights pay without having taken that decision has not avoided it. It has delegated it to the evaluation panel.
Nothing here has been tested against a live PPN 026 procurement, because there have not been any; the model applies to procurements commenced on or after 1 January 2027, so this is where our thinking has got to rather than a method.¹ We would be glad to hear from any organisation already doing this arithmetic locally.
1. Take the budget decision first, and take it together. Before selecting criteria, work out what the contracted workforce currently costs, what the terms the authority wants would cost, and where the difference would come from. That conversation needs the service manager, the commercial team and whoever owns the budget in the same room, not a procurement template.
2. Weight pay only where the money is real. An authority that will fund better terms can weight MAC 1c and mean it. Meaning it includes intending the more expensive bid to win; a service manager and commercial team surprised by that outcome usually end up descoping the contract or reprocuring it. One that will not should weight the access criteria (1a, 2a and 2c) honestly instead, and carry pay through the specification or as one of the KPIs section 52 of the Procurement Act 2023 already requires, which keeps it enforceable without pretending the marks will buy it.
3. Treat genuinely new work differently. Where a contract is new, scope has expanded, or a supplier is building capacity it did not have, new posts can be created on the promised terms rather than inherited ones, and better jobs and wider access stop competing with each other. On a like-for-like recompetition that room is mostly not there, and the budget decision is the honest place to start.
How the six criteria behave once the budget is settled, and what the marks are actually worth at evaluation, are the subjects of the next two articles in this series.