Section 136 Equality Act 2010
"(2) If there are facts from which the court could decide, in the absence of any other explanation, that a person (A) contravened the provision concerned, the court must hold that the contravention occurred. (3) But subsection (2) does not apply if A shows that A did not contravene the provision."
This applies to every Equality Act claim — direct, indirect, harassment, victimisation, s.15 and reasonable adjustments.In plain English, the case runs in two stages:
Stage 1 — You establish prima facie facts
You must prove facts from which the tribunal could conclude, absent an explanation, that discrimination occurred. This is not proof of discrimination itself — it is evidence of a difference in treatment plus something suggesting the prohibited ground: a comparator treated better, a suspicious pattern, a comment, a timing sequence, statistical evidence.
Stage 2 — The burden shifts to the employer
Once stage 1 is made out, the employer must prove, on the balance of probabilities, that the treatment was in no sense whatsoever because of the protected characteristic. A bare assertion is not enough — they need a credible, evidenced, innocent explanation.
Claimants lose when they argue their case in the air — "it was obviously discrimination" — instead of proving concrete facts: dates, comparators, documents, witnesses. Everything on this page is about making stage 1 solid.
The three cases you should know
Igen Ltd v Wong [2005] EWCA Civ 142
The definitive guidance on the two-stage test (codifying the earlier Barton v Investec approach). Key points tribunals apply daily: at stage 1 the tribunal must assume there is no adequate explanation and ask what the facts point to; at stage 2 the employer's explanation must be adequate and proved; unreasonable treatment alone is not automatically discrimination — but an employer who behaves unreasonably gives tribunals material from which inferences can be drawn.
Efobi v Royal Mail Group Ltd [2021] UKSC 33
Refined Igen for the post-2010 statute: s.136 does not impose a formal "burden" on the claimant at stage 1 — the tribunal must consider all the evidence, from all sources (including the employer's own witnesses and documents), when deciding whether prima facie facts exist. Good news for litigants in person: hostile witness evidence and disclosure documents count towards your stage 1.
Hewage v Grampian Health Board [2012] UKSC 37
The tribunal must not only consider whether discrimination could be inferred — it must be satisfied, on the whole case, that it actually occurred. The "could decide" test is not a rubber stamp: weak stage 1 evidence that is comprehensively rebutted will still fail. Lesson: treat stage 1 as the floor, not the ceiling, of your evidence gathering.
What actually shifts the burden
| Evidence type | Example | Strength |
|---|---|---|
| Actual comparator | A named colleague without your characteristic, same role, treated measurably better for the same conduct | Strongest for direct discrimination |
| Hypothetical comparator | Evidence of how the employer treated others generally, used to construct how a comparator would have been treated | Strong, needs pattern evidence |
| Comments and remarks | Dated, recorded remarks linking the characteristic to the decision ("we need fresh blood", "you're emotional") | Strong if contemporaneous |
| Timing sequence | Disclosure of pregnancy → adverse action within weeks; clean record → sudden "capability" process | Corroborative — rarely enough alone |
| Statistics | Promotion or dismissal rates by sex, race, age across the organisation | Powerful for indirect; requires data |
| Procedural anomalies | Policy not followed for you but followed for others; investigation that ignored your witnesses | Corroborative — feeds the inference |
Use the comparator evidence worksheet to map each incident against a comparator, a date, a witness and a document — and use a SAR to obtain the internal documents that fill the gaps.
The questions procedure
Since 2014 the statutory discrimination questionnaire is gone, but its replacement is still powerful: write to the employer before filing your claim with focused questions — Who decided? On what criteria? What documents record the decision? How were others in the same situation treated?
- The employer is not obliged to answer — but the tribunal may draw adverse inferences from evasive or absent replies (this survives the repeal; ACAS guidance confirms it).
- Answers bind the employer later: an explanation given in February that changes by the hearing is stage-2 dynamite.
- Keep questions factual and specific; rhetorical questions get ignored.
Evidence gathering does not extend the three-month-less-one-day limit. Send the SAR and the questions early, and start ACAS early conciliation in parallel. The Tribunal Guide explains how the conciliation pause works.
Putting it into practice: six rules the appeal courts repeat
The principles above come from dozens of appeal decisions. Distilled, they give you six rules for building a discrimination claim that survives.
- Tick the statutory boxes first. Direct discrimination is section 13 of the Equality Act 2010: less favourable treatment because of a protected characteristic. The tribunal's power to hear the claim comes from s.120, and the remedies — a declaration, compensation, and recommendations — from s.124. Frame your claim in those terms from the outset.
- Establish your comparator. The comparator question is "at the heart of the case." Use a real colleague if one exists — but remember Wakeman v Quick: a hypothetical comparator (how someone without your characteristic would have been treated) is expressly permissible.
- Plead every act, specifically. The EAT overturned a race discrimination finding because the tribunal relied on acts that had never been pleaded. List each incident separately: date, what happened, who was involved, and why it points to discrimination.
- Build the inference. Direct evidence is rare — the EAT notes it is "the covert rather than the overt act" that usually provides evidence of discrimination. Tribunals can infer discrimination from a pattern: statistics, timing, inconsistencies in the employer's story, and how others were treated.
- Use stereotyping as evidence. Comments revealing stereotypical assumptions about people with your characteristic can support a finding of direct discrimination — provided they are tied to the correct comparator analysis, not left to stand alone.
- Anticipate their explanation. Once you establish the basic facts, the burden shifts — and the employer must prove, on the balance of probabilities, a non-discriminatory explanation. Rebut it in advance: an employer who can only show their process was fair has not answered the discrimination question (Zafar v Glasgow City Council).
Tribunals must apply the established framework and give proper reasons for each issue (Meek v City of Birmingham DC). Where they fail, the EAT will send the case back. And at case management, a tribunal must not cut down the complaints you are allowed to pursue — restricting your claims risks denying access to justice (Dr Tarn v Hughes). Our Key Cases library covers the landmark authorities in plain English.