Social value is scored. It is rarely measured

The duty that started all of this is a duty to think. Section 1 of the Public Services (Social Value) Act 2012, in force since 31 January 2013, requires a contracting authority to consider how what it proposes to procure might improve the economic, social and environmental well-being of the relevant area, and how it might act during the procurement itself to secure that improvement.

Consider. Not achieve, not measure, not report. Fourteen years of practice have been built on a verb that cannot be failed.

The weighting keeps going up

What has changed is the price of thinking. PPN 002, which applied to central government procurements commenced under the Procurement Act from 1 October 2025, set a 10% minimum weighting of the total score for social value, or an equivalent measurement, and required that commitments made during the procurement be reflected in the contract as terms, key performance indicators or performance indicators.

PPN 026 goes further. From 1 January 2027, covered procurements worth £1 million or more attract a minimum 10% weighting, and those at £5 million or above attract a minimum 20%. It also narrows the definition, describing social value for central government contracts as taking account of how a supplier will work for our communities to provide good British jobs, skills and opportunities in every postcode.

That narrowing is the more consequential half. The earlier models spread across employment, supply chain resilience, environmental outcomes, equal opportunity and community wellbeing. The new one concentrates on jobs, skills and opportunities. Respondents to the consultation behind it asked for environmental outcomes, health and wellbeing, and diversity and inclusion to stay in scope. The government response, presented to Parliament on 26 March 2026 after 811 survey responses, kept the tighter focus, and reported that respondents also warned against pushing the weighting higher on the grounds that it would disadvantage smaller bidders.

Twenty per cent of what

The number that gets quoted is 20%. The number that decides awards is the spread.

Weighting determines how much of the score social value can move. It does not determine how much it does move, and on most competitions it moves very little, because every serious bidder employs people who write social value responses for a living and the resulting submissions cluster. Where the field lands within a few marks of itself, the criterion produces a few points of differentiation regardless of whether it carries 10% or 20% of the total.

How tight that cluster actually is nobody has published. No authority releases the distribution of social value scores across a set of competitions, so the size of the effect is asserted by bidders and evaluators from their own experience and cannot be checked by anyone else. That is a gap somebody could close with the notices that already exist.

There are two ways that changes. Either evaluators start using the full range of the scale, which means publishing scores low enough to be challenged, or the commitments become contractually binding at a level that makes bidders price them differently. The second is happening quietly, and it is the part of this reform worth watching.

The measurement machinery already exists

This is the fact that reframes the argument. Section 52 of the Procurement Act 2023 obliges an authority to fix at least three key performance indicators for any public contract it expects to be worth more than £5 million, and to publish them. The provision took effect on 24 February 2025. Subsection (6) puts four categories outside it: frameworks, concessions, light touch contracts, and utilities contracts awarded by a private utility. Note the qualifier on the last of those. A utilities contract awarded by a public authority stays in scope. Contracts whose performance cannot sensibly be judged against indicators are also excepted.

Under section 71, at least once in every twelve months during the life-cycle of the contract, and on termination, the authority must assess performance against those indicators and publish the specified information about that assessment.

Read those two sections together with the PPN requirement that social value commitments become contract terms or KPIs, and the objection that social value cannot be measured collapses. A commitment made in a bid becomes a KPI, the KPI is published at award, and performance against it is assessed annually and published. The machinery is on the statute book and it commenced eighteen months ago.

So the real question is who reads it

Nobody is required to. That is the honest position.

An assessment published under section 71 goes into the notice system, where it joins award notices, contract change notices and everything else. There is no obligation on any authority to act on a poor assessment beyond its ordinary contract management, no obligation on any other authority to look at a supplier’s published record before shortlisting, and no central body producing a view across the whole of it.

This is a countable question rather than an arguable one, and the count is the interesting piece of work in this field. Take the contracts above £5 million awarded since 24 February 2025, look for the KPIs published under section 52, then look for whether an assessment has been published against them where twelve months have elapsed. Nothing in that method requires privileged access or a freedom of information request. It requires patience with the notice system, which is precisely why it has not been done.

Any claim about whether social value is delivered, from a supplier or a critic, that does not rest on that count is an opinion wearing a number.

The measurement problem that is real

There is one objection to social value scoring that survives all of this, and it deserves stating properly rather than being used as a general excuse.

Attribution. A supplier commits to creating twenty apprenticeships during a five-year contract. It creates twenty apprenticeships. Whether any of them exist because of the contract, or whether the firm would have recruited at that level anyway and simply allocated the count to whichever bid asked for it, is not knowable from the outside and frequently not from the inside either. The same applies to local spend, to volunteering hours and to almost every metric in the standard models.

That is a genuine limit and it argues for a specific response. Prefer commitments that are additional by construction. A payment term is additional: the supplier either pays in thirty days or it does not, and the counterfactual does not arise. A subcontracting target is close to additional. An apprenticeship count is not, and the models still lean on it because it is easy to state in a bid.

The move from scored aspiration to verified condition is therefore not just an enforcement preference. It is the only version of social value that is measurable at all.

The pattern in the reform, and what it is really for

Two other changes announced in the March 2026 response give the game away about the policy’s direction. On the largest contracts, a supplier that cannot evidence prompt invoice payment is to be shut out altogether, and the consultation’s proposed 60-day yardstick met a push from respondents for 30. Separately, authorities whose annual procurement passes £100 million are to publish three-year goals for the share of their spending reaching smaller firms and social enterprises, reporting each year on progress.

Both are social value provisions in everything but name, and both are structured completely differently from the scored kind. They are thresholds and published targets rather than criteria. You either pay in thirty days or you are out. You either hit the SME target or the miss is on the record.

That contrast is the answer to the question in the title. Social value as an award criterion has fourteen years of practice and very little published evidence of delivered outcomes. Social value as an exclusion ground, a mandatory KPI or a published target has almost no practice and produces evidence automatically. The reforms are gradually moving weight from the first form to the second, which suggests the people who designed the original model have drawn the same conclusion.

The broader buying picture is in what the Procurement Act changed for suppliers, and the rest of this coverage sits under public sector technology.