HMCTS : Employment Appeal Tribunal Rule 3(10) – The Sift Was Not The Case Before Lazy Tayler

By /

Thompson’s appeal against Cummins had already been rejected at the initial sift. By the time it reached HHJ Tayler under Rule 3(10), however, I had responded to that rejection, expressly identified errors of law and put a completely fucking different evidential bundle before the Employment Appeal Tribunal. The EAT’s own Practice Direction says a Rule 3(10) judge considers the matter “afresh”. What I got looked rather more like somebody reaching backwards for the original rejection and refusing to wrestle properly with the case now sitting on the desk. That raises an uglier question: how willing is the EAT to correct judicial fuckery when doing so means admitting that another judge may have seriously fucked up?

The Sift Was Not The Case Before Tayler

This really should not require a legal séance.

The original sift considered one state of the appeal. After that rejection, I went back through the case, addressed what had been identified as deficient, sharpened the legal arguments and expressly advanced errors of law. Most importantly, I assembled a Rule 3(10) evidential bundle to demonstrate why those errors were arguable.

By the time HHJ Tayler dealt with the matter, the landscape had changed.

The legal argument had developed. The evidence before the EAT had changed. The material relied upon to demonstrate the alleged errors had changed. I was not simply standing at the same counter waving the same fucking receipt and asking another judge to look at it.

I had answered the sift.

That matters because Rule 3(10) is not supposed to be a commemorative service for the original rejection. According to the EAT’s own Practice Direction, the hearing gives an appellant an opportunity to persuade a judge that reasonable grounds exist for bringing some or all of the appeal.

Then come the four words that make this interesting.

The judge considers the matter afresh.

Afresh.

Not embalmed.

Not reheated.

Not photocopied from the previous judicial opinion with a fresh signature at the bottom.

Afresh.

The EAT Wrote The Fucking Rulebook

The same Practice Direction does allow a Rule 3(10) judge to adopt some or all of the sift judge’s reasoning. There is nothing inherently improper about two judges reaching the same conclusion.

But read the whole fucking paragraph.

First, the matter is considered afresh. Then the Practice Direction says the judge may adopt the sift reasoning, particularly where the appellant has not explained why the Rule 3(7) opinion is wrong.

That distinction could hardly be more important here.

I had explained why I said it was wrong.

I had identified the alleged legal errors. I had developed the case beyond the material originally sifted. I had also put a completely different bundle before HHJ Tayler, containing material I said demonstrated those errors.

The EAT’s own Practice Direction expressly provides for supplementary documents at a Rule 3(10) hearing where they are necessary to the argument. It requires those documents to be identified and their necessity explained in a skeleton argument. It even contemplates amendments to a Notice of Appeal being permitted at the Rule 3(10) hearing itself.

In other words, the EAT’s own machinery recognises something painfully obvious: the case presented at Rule 3(10) may not be frozen in exactly the form that generated the sift decision.

So what happens when the appellant actually develops it?

That is where my confidence in this process fell through the fucking floor.

The Bundle Was Not Courtroom Wallpaper

My Rule 3(10) bundle existed for a reason.

It was not decorative paper surrounding a judgment the EAT had already decided it liked. It was not there to make the PDF heavier or give everybody something impressive to flick through while repeating the original answer.

The evidence was there because I was identifying legal errors and showing the material upon which those arguments depended.

If my arguments were hopeless, kill them.

If the documents did not support the alleged errors, explain why.

If I had misunderstood the law, identify the misunderstanding and put the point in the ground.

That is appellate scrutiny. It does not require the judge to agree with me. It requires the judge to engage with the actual fucking argument.

Instead, the reasoning I received felt alien to the developed case I had put before the EAT. The original sift remained the gravitational centre while the intervening legal arguments and evidential work seemed to orbit somewhere outside meaningful consideration.

Agreement with an earlier decision is the possible result of fresh analysis.

It is not the analysis itself.

A Different Case Cannot Be Answered By The Old Receipt

The later Court of Appeal order makes the procedural evolution visible in black and white.

Lord Justice Nugee records that one EAT ground had contended that dismissal was on capability grounds arising directly from disability. He then records that, at the Rule 3(10) stage, I argued that the reason for dismissal was conduct and/or capability and that this was arguably disability-linked behaviour.

Whatever anybody thinks of the merits, that description itself demonstrates movement in the argument.

The Court of Appeal ultimately rejected my attempt to take the matter further. Nugee agreed with Tayler on the Section 15 issue and also rejected my complaint about the brevity of the Rule 3(10) reasons.

Fine.

TCAP is not hiding the adverse decision in a cupboard.

But brevity was never the real fucking disease.

Engagement was.

A judgment can be ten pages long and still miss the point. It can be ten paragraphs long and nail it perfectly. Word count is irrelevant if the reasoning never adequately confronts the developed legal error and evidential case that the appellant says transformed the issue after sift.

The question remains brutally simple.

What was my Rule 3(10) bundle for?

Is The EAT Scared Of Correcting Judicial Fuckery?

This is where the problem stops being merely technical.

Judges make mistakes. Everybody capable of using cutlery makes mistakes. An appellate structure exists precisely because judicial decisions are not sacred objects delivered from a mountain.

Yet correcting a serious judicial error requires something institutionally uncomfortable.

Another judge has to say a judge below got it wrong.

Sometimes badly wrong.

Sometimes in a way that means pulling at one thread and discovering that half the fucking jumper comes with it.

Looking at my experience, I am entitled to ask whether the Employment Appeal Tribunal is genuinely comfortable doing that when the alleged error strikes deeply into the reasoning below. Because the alternative explanation for what happened at Rule 3(10) is increasingly difficult to ignore.

Maybe the EAT simply got my appeal wrong.

Or maybe appellate caution becomes something more defensive when properly examining the alleged error might expose judicial fuckery underneath.

That is not an allegation of some smoky-room conspiracy. Nothing so glamorous is required. Institutions protect themselves perfectly well through habit, deference and an instinctive preference for yesterday’s answer.

The file arrives carrying a judgment.

The judgment carries authority.

The litigant carries the burden.

Then every benefit of institutional gravity rolls downhill towards preserving what has already been decided.

That is precisely why “afresh” fucking matters.

The Statistics Are Not A Comfort Blanket

Official Ministry of Justice figures make the terrain interesting too.

For 2024/25, 31 per cent of employee appeals dealt with at an EAT preliminary hearing were dismissed, compared with 15 per cent of employer appeals. Among appeals that reached a final EAT hearing, 57 per cent of employee appeals were dismissed compared with 41 per cent brought by employers.

Those are not Rule 3(10)-specific statistics, and HMCTS itself says the EAT figures are under data-quality review.

Even so, they hardly describe a landscape tilted towards the individual employee arriving to say a legal wrong needs correcting.

More importantly, statistics cannot answer an individual legal error.

A low success rate does not prove the system is excellent. Nor does the fact that most challenges fail magically prove the judgments being challenged were sound. Numbers can tell us what happens to appeals. They cannot tell us whether the losing appellant’s argument was properly confronted.

There is also something courts and tribunals should remember about the people who get this far.

Most ordinary people do not wake up craving appellate litigation. They do not spend months reading judgments, identifying legal tests, constructing arguments, preparing bundles and taking disputes towards a specialist appellate tribunal because the fucking television schedule looked weak.

They do it because they believe something seriously wrong happened to them.

That belief does not make them right.

But the fact that relatively few succeed does not make them wrong either.

When Rejection Starts Feeding Rejection

There is a dangerous circularity here.

The original judgment carries authority because a judge made it. An appeal is difficult because only errors of law will do. The sift then filters the appeal through that narrow gateway. If it fails, the appellant gets a Rule 3(10) route and must explain why the sift was wrong.

So far, fine.

But if the later judge approaches the developed challenge through the gravitational pull of the earlier rejection, the process begins eating its own paperwork.

The judgment was probably right because it survived scrutiny.

The sift was probably right because the judgment looked defensible.

The Rule 3(10) rejection was probably right because another judge had already sifted it.

Then the low number of successful appeals becomes another comforting institutional statistic suggesting the machinery must be working beautifully.

A closed loop with a wig on.

That is why the Rule 3(10) obligation to consider the matter afresh is not procedural garnish. It is supposed to break that loop.

Otherwise an appellant can correct the formulation, identify the alleged errors, assemble the evidence, comply with the bundle requirements and still find himself arguing against the ghost of a case he no longer fucking presented.

Nugee Closed The Next Door

On 17 July 2026, Lord Justice Nugee refused permission for me to appeal the Rule 3(10) outcome further.

His order says the proposed appeal had no real prospect of success and no other compelling reason to be heard. On the Section 15 issue, he agreed with HHJ Tayler. On the late timing evidence, he also agreed that the point was unarguable. The Court of Appeal door therefore closed.

That is the procedural outcome.

It does not require TCAP to pretend the reasoning suddenly makes sense.

Nor does the finality of a judicial decision convert criticism of it into heresy. Courts publish judgments precisely because judicial reasoning is capable of being read, tested, criticised and compared against the record.

And my criticism is now narrower than ever.

I am not saying HHJ Tayler had to agree with me.

I am saying that the Rule 3(10) process demanded meaningful engagement with the case that actually existed by the time it reached him.

That case was not the sift.

The Court Had One Fucking Job

By Rule 3(10), I had done what the process demanded of me.

I identified alleged errors of law. I developed the arguments. I assembled the evidence. I put a different bundle before the EAT and explained why the original rejection did not answer the case I was now advancing.

The EAT was perfectly entitled to destroy every point.

What it needed to do was destroy those points.

Instead, I came away with reasoning that felt disconnected from the legal errors and evidential case I had actually put before it, while the earlier sift continued to cast the decisive shadow.

That is where this moves beyond ordinary disappointment at losing.

When somebody goes to the lengths required to put a developed legal argument before an appellate tribunal, the minimum legitimate expectation is not victory. It is recognition. The resulting reasoning should bear some fucking resemblance to the case that went through the door.

Mine did not.

So I am left with a question that the Employment Appeal Tribunal may find considerably less comfortable than another bundle.

Is the EAT genuinely prepared to correct judicial fuckery when the evidence demands it, or is preserving the judicial decision already on file simply the path of least institutional resistance?

Because when an appellate tribunal says it will consider a matter afresh, receives the developed legal errors and the completely different fucking bundle, then produces reasoning that remains alien to what was actually put before it, we are no longer talking about somebody merely being angry because he lost.

We are talking about an appellate safeguard that appears to have looked straight at the safeguard and stepped around it.

That borders on scandal.

Lee Thompson – Founder, The Cummins Accountability Project

Sources

Scroll to Top