The Duty Before the Decision: Evidencing Equality Analysis in Commissioning

The Duty Attaches to the Decision, Not the Document

The Public Sector Equality Duty is often treated as a paper exercise appended to a committee report. In commissioning, that misunderstanding is costly. Section 149 of the Equality Act 2010 requires a public authority, in the exercise of its functions, to have due regard to three needs: eliminating discrimination, harassment and victimisation; advancing equality of opportunity between those who share a protected characteristic and those who do not; and fostering good relations between different groups.

The duty is not a duty to reach a particular outcome. It is a duty about how the decision is taken. The authority must understand the duty, have before it sufficient information to assess the equality implications of what is proposed, consider that information with an open mind, and do so before the decision is made — not afterwards in justification of it. That much was settled in R (Brown) v Secretary of State for Work and Pensions and reinforced in R (Bracking) v Secretary of State for Work and Pensions, where the Court of Appeal emphasised that the regard must be exercised in substance, with rigour and with an open mind, and that a decision-maker cannot rely on an analysis they have never seen.

Commissioning sits squarely within this territory. A decision to outsource a service, to consolidate two contracts, to change eligibility criteria, to award on cost-weighted criteria or to extend an existing arrangement will redistribute access, quality and employment. The duty is engaged at each of those points, and it is engaged again when the contract is varied or renewed.

Where Commissioning Evidence Usually Fails

In our experience, four weaknesses recur.

The first is timing. An equality impact assessment (EqIA) produced after the evaluation panel has scored the bids, or signed at the same meeting at which the award is approved, is evidence of process, not of regard. The decision-maker has not had the opportunity to shape the proposal in response to it. Where options have already been closed off, the analysis cannot do the work the duty requires.

The second is generality. Statements that a service will be delivered "in accordance with the Equality Act" or that "no adverse impact is anticipated" are assertions. They do not identify who is affected, how, or on what evidence. A court will look for a conscious engagement with the likely effects of the specific proposal, including any differential effect on disabled service users, on carers, on older people or on groups with particular religious or cultural needs.

The third is disaggregation from the decision. A well-written EqIA that sits in an appendix while the evaluation model rewards price at 70 per cent, and the specification is silent on accessibility, has had no influence on the outcome. The duty is about the decision, and the decision is the combination of specification, award criteria, weightings and contract terms.

The fourth is cumulative effect. Individual contract decisions may each appear benign while, taken together across a place or a service pathway, they reduce provision for a particular group. Authorities that commission across several lots, or across several years, need to consider the aggregate position rather than assess each procurement in isolation. The same discipline applies to a series of modifications that, taken singly, look modest.

Anchoring the Analysis in the Business Case

Equality analysis is most effective when it is written into the business case rather than bolted onto the report. At the options stage, the authority should set out who currently uses the service, who does not, and why. Sources will vary, but typically include equality monitoring data, the joint strategic needs assessment, census data, complaints and casework themes, engagement with representative groups and the views of frontline staff.

The analysis should be option-specific. It is not enough to assess the preferred option; the discarded options should also be assessed, because the decision includes the choice not to pursue them. Where mitigation is proposed — an accessibility requirement in the specification, a transition period, a hardship provision — the analysis should say what it is intended to achieve, who owns it, and how its effect will be checked.

Residual impact should be stated plainly. An honest assessment that a change is likely to have a modest adverse effect on one group, mitigated by specific measures and monitored through defined indicators, is a stronger basis for a lawful decision than an implausible claim of neutrality.

Carrying the Analysis into the Procurement

Due regard must survive the journey from business case to contract. That means the conclusions of the analysis should be visible in the specification, in the conditions of participation, in the social value or quality questions, and in the weightings applied. Where accessibility, workforce continuity or service-user support matters, the evaluation model should make it scoreable, and the moderator or evaluation report should record how the responses were assessed against it.

Authorities operating under the Procurement Act 2023, in force since February 2025, have a clearer statutory framework for award criteria and for assessing the most advantageous tender, but the equality analysis duty is unchanged and sits alongside it. The two regimes should be read together, not sequentially.

Award, Modification and the Record

At the point of award, the record should show that the decision-maker saw the analysis, understood its conclusions, and took them into account. If the outcome departs from the analysis — for example, because the winning bid proposed a different service model — the reasoned record should say so and explain what mitigations were secured. The standstill period is not the occasion to construct that reasoning retrospectively.

The same discipline attaches to modifications, extensions and renewals. A contract change that alters eligibility, location, hours or channel of access is a fresh decision with fresh equality implications, even where it does not require a new procurement.

Contract Management as Ongoing Evidence

Due regard does not end at signature. Where the analysis identified risks, the contract should carry proportionate monitoring: service-user data disaggregated in a way that permits meaningful review, reporting against agreed indicators, and defined review points. A contract management file that shows mitigation was delivered, and that adjustments were made where it was not, is itself evidence of a continuing conscious approach.

A Short Assurance Test

Before a commissioning decision is taken, four questions are worth asking. Was the analysis available to the decision-maker before the recommendation was formed? Does it address the options actually available and the cumulative effect of what is proposed? Did it change anything — the specification, the criteria, the weightings, the mitigations or the timetable? And would a reader outside the organisation be able to follow the reasoning from evidence to conclusion?

If the answer to any of those is no, the issue is not the length of the assessment. It is whether due regard has been exercised at all.