Brexit and the waning days of the United Kingdom

[Sheilbh]

So Policy Exchange (fairly influential right-wing think tank) have published their paper on judicial reform and, surprisingly, I'm seeing some liberal/left barristers actually being kind-of convinced by it.

Basically their proposal is to abolish the Supreme Court as a standing institution. It wouldn't get rid of the third appellate tier but instead of having a fixed Supreme Court, it would draw its membership on a case by case basis from the Courts of Appeal across the UK.

Their argument isn't that the name or physical location matters but the hierarchy creates a sort-of judicial hubris and celebrity judges. In addition the report notes that the House of Lords and Supreme Court haven't been full of judges who are clearly significantly better than you get at the Courts of Appeal level - there are some very good judges in the Courts of Appeal and some weaker ones on the Supreme Court. So basically instead of a fixed Supreme Court you have, in effect an Upper Court of Appeal or Final Court of Appeal.

In addition to removing the sort of celebrity judge risk it would also probably further insulate the judiciary from attack if politicians couldn't have their bete noire judge (eg Baroness Hale or Lord Sumption), instead there'd just be a rotation of the Courts of Appeal judges. So far we've been lucky in not having a particularly politicised judiciary, but you only need to look at other common law countries to see the risk.

So not as bad as feared and actually, maybe a sensible proposal?

[Tamas]

I don't know nearly enough about how this works but the need to assemble a separate gathering of Court of Appeals judges on a case by case basis would not diminish the chances of cases actually going to supreme court level? Sounds like if you are defeated in the Court of Appeals you could appeal to the Court of Appeals to have another look by different judges and then that would be it. Especially with us out of the EU.

[Richard Hakluyt]

The supreme court itself is an innovation and one that I do not particularly like. Much as i enjoyed Baroness Hale sticking it to the government far too much of that joy was partisan. The judiciary has been relatively non-political as you say, this change would reinforce that.

The concern would be that the rotating judges court would be too weak to hold a badly-behaved government to account.

[Sheilbh]

I don't know nearly enough about how this works but the need to assemble a separate gathering of Court of Appeals judges on a case by case basis would not diminish the chances of cases actually going to supreme court level? Sounds like if you are defeated in the Court of Appeals you could appeal to the Court of Appeals to have another look by different judges and then that would be it. Especially with us out of the EU.

I think the basic proposal is that after two years of sitting as a judge in one of the Courts of Appeal, judges would be given a new title and sit on a panel of judges. Judges on that panel would still sit in normal Court of Appeal cases, but could also be selected to sit on cases to the "Upper/Final Court of Appeal".

The author suggests because this will increase the workload expanding the English Court of Appeal by another 9-10 judges and similar changes to the Courts of Appeal in Scotland etc. Obviously current members of the Supreme Court would immediately go into the panel. You could also in effect have ex officio members who are retired judges (under the mandatory retirement age) if there's not enough people on the panel of the "Upper/Final Court of Appeal". I think there's sort of an almost philosophical point of what is the purpose of a final appellate court (which the EU courts aren't, incidentally) - the current model is basically one of error correction, what he's proposing is more along the lines of a sort-of peer review.

It's an interesting proposal by a law professor and QC which seems to slightly disappoint the think-tank writers who were all very excited (as you'd expect) at the prospect of just somehow reverting to the old House of Lords model. He's also very dismissive of the ideas that the think tank had about the new name or location causing the Supreme Court to act differently - he doesn't think the Supreme Court is acting that differently from the old House of Lords but thinks they both share similar issues.

Joshua Rozenberg is not particularly convinced - but as I say a lot less Jacobin and a lot more interesting than I would have expected:

Let's abolish the UK Supreme Court!
But what would we put in its place?
Joshua Rozenberg
21 hr   

Let's abolish the UK Supreme Court! So says Professor Derrick Wyatt QC, a distinguished former academic and practitioner, in a paper for the Policy Exchange judicial power project.

It's fair to say that the title of his paper speaks of reform rather than abolition; that his proposals are avowedly tentative; and that he would change the Supreme Court's name and composition rather than do away with a final tier of appeal in its entirety. But this is still a pretty radical proposal.

Supporters of the judicial power project believe the Supreme Court (pictured) is "too activist". That's not my view — see my recent book Enemies of the People? — and it's not a view that Wyatt fully supports either. As he says, the same criticism could be levelled at the appellate committee of the House of Lords, which the Supreme Court replaced in 2009.

But Wyatt thinks the risk of excessive judicial activism at the final stage of the UK appellate process could be "tempered" by transferring the Supreme Court's functions to panels of five or more senior judges assigned on a case-by-case basis from judges in the existing appellate courts — the Court of Appeal of England and Wales, the Court of Appeal of Northern Ireland, and the Inner House of the Court of Session.

These judges would spend most of their time in their respective courts of appeal. But they would move to the Supreme Court's current building in Parliament Square (or elsewhere in the UK) when hearing appeals from the courts in which they normally sat. As Wyatt explains,

    the result of the above change would be to broaden the judicial base of the final court of appeal, and to ensure that all judges sitting at the final stage of appeal were also judges deciding cases at the level below that final stage, and so remained accustomed to the discipline of writing judgments which would be subject to the scrutiny of their fellow judges. This would I believe militate against the final court of appeal being, or being seen to be, a judicial policy making centre independent of government or parliament, with perhaps a corresponding sense of entitlement to rebalance the constitutional order of things as it thought fit.

From a practical point of view, Wyatt's proposal would increase diversity in what might be called the UK Final Court of Appeal. On the other hand, it would thin out existing appeal courts that are already under pressure. Wyatt has various transitional provisions to address this.

But the fundamental problem with his proposal is one touched upon by Professor Richard Ekins in an afterword: "how — and by whom — panels would be selected for particular cases". Ekins inclines to the view that panels should be "formed simply on the grounds of availability, taking for granted that appellate judges are or should be capable of deciding any case before them". But, as he seems to acknowledge, random selection is unsustainable.

In appeals from Northern Ireland — and, increasingly, from Wales — it is desirable that there should be at least one judge who has previously sat in those jurisdictions. In appeals from Scotland, it is essential. The Supreme Court tries to ensure that panels include a judge with experience of the issues under appeal. A family appeal, for example, should not be heard by a court of five family specialists — but neither should it be decided by a court of none.

As Ekins accepts, selecting the judicial panel is an "important and difficult question". He does not explore the difficulties — but they are not hard to see.

Let us accept that, in each case, some members of an appeal court are more suitable than others to hear an appeal from their colleagues. They may be more senior. They may have specialist knowledge of crime, or trusts or whatever the issue may be. The three appeal courts currently select panels, as far as possible, to reflect those considerations. If a case is particularly difficult or important, the chief justice may sit with senior colleagues.

Who would then be sent off to the Final Court of Appeal to hear an appeal against their colleagues' decision? Non-specialist judges? Those with less seniority?

That problem could be avoided by holding back the most appropriate judges, leaving them free to hear the final appeal. But what if there isn't one? A case would then have been decided by a less appropriate panel for no good reason. The parties might then need to appeal — at great expense to all concerned — just to make sure their case was heard by the judges who would have heard it under the current system.

Selecting judges to sit on individual cases is an important judicial function. Who would decide which judges should hear each case in the Final Court of Appeal? The three chief justices of the United Kingdom? Two of them would know little about the candidates. Each would have to decide whether to sit themselves. And their choices could be seen as affecting the likely outcomes.

Although I do not regard the Supreme Court as too activist, there is no doubt that some judges — at all levels — are more activist than others. They may be activist by temperament. In others, it may depend on the case: Lord Reed, president of the Supreme Court, told me he was more inclined to develop the law in some areas than others. It's because the outcome of a case may depend on the choice of judge that 11 members of the Supreme Court — the maximum — considered the constitutional challenges brought by Gina Miller in 2016 and 2019.

But that would not be possible under Wyatt's plans: there would be too many people on the bench. Somebody — and it could only be a very senior judge — would have to choose the judges who would hear each appeal. That judge could choose colleagues who were likely to take an activist approach — precisely the problem the judicial power project seeks to challenge.

And then there are the practical problems. It's one thing to be overturned by a higher court. It would be much more difficult to be overturned by a junior colleague who you are sitting with the following week. Perhaps that's the point: in his introduction to Wyatt's paper, Lord Thomas of Cwmgiedd, the former lord chief justice, asks whether "the prospect of judgments being scrutinised by fellow judges exercising an appellate function encourage self-restraint and improve accountability". It might; but do we really want judges to neuter themselves?

I have enjoyed thinking about Wyatt's ideas. His proposals deserve serious debate at academic seminars. But I don't think the 12 justices of the Supreme Court will be looking for new jobs any time soon.

[Tamas]

Yeah these are very good points. It did sound like this would turn into a peer review of Court of Appeal decisions by the colleagues of the judge making the original appeal ruling. How could we realistically expect the "review" to be independent of, well, office politics, basically.

Also, as stated, who would do the mixing and matching of judges?

[Sheilbh]

The supreme court itself is an innovation and one that I do not particularly like. Much as i enjoyed Baroness Hale sticking it to the government far too much of that joy was partisan. The judiciary has been relatively non-political as you say, this change would reinforce that.

Yeah and I think one of the strengths of this proposal is he doesn't act, like conservatives often do after the Brexit decisions, like the Supreme Court is acting in a wildly different way than the House of Lords. He says there's no real difference which I think is right.

I think one upside of the Supreme Court (and this wouldn't change) is the optics. I always remember one of the Law Lords commenting that a reason he moved to supporting a separate body/building of the Supreme Court was giving the judgment in the Pinochet case in the House of Lords with Baroness Thatcher sat two rows behind you. The Law Lords were institutionally independent and I don't think were influenced by stuff like that, but that judgment was being shown around the world and it looked very strange/problematic.

The concern would be that the rotating judges court would be too weak to hold a badly-behaved government to account.

I wouldn't worry about that too much. You'd have the same personal issues of it being a job for life and also the final job a judge will hold which I think increase their independence, even sub-consciously. But also I don't get the sense that the Court of Appeal at the minute is particularly cowed by government.

[Gups]

I agree - no real difference between the HoL and UKSC except the facilities are much, much better.  I had the second to last case the HoL heard and it was a complete pain in the arse getting through security with bundles and ensuring clients could find somewhere to have lunch while at the same time discussing the morning's proceedings with the QC (with their own dining facilities).

Main probllem with the Court of Appeal is the waiting list. I had a case given permission this JUne and it has an indicative hearing date for November 2021. We still had to put the skeleton in four weeks after permission was granted though.

[Sheilbh]

Also, as stated, who would do the mixing and matching of judges?

Agreed. But that is something that already exists in our system. So I think the Court of Appeal in England sit in groups of 3 but I don't know who selects the judges to hear which cases.

Similarly the Supreme Court has 12 judges but they sit in panels to hear cases. I think most cases get 5 justices and then ones that are particularly important or might involve overturning precedent or reconciling different precedents have a larger panel - so the Brexit cases got a full panel of 11 judges because they were very important and high profile. But I'm not sure how it works - I assume the President picks which judges are on which cases. Which is certainly more powerful than the Chief Justice in the US Supreme Court who will just pick who gets to write the opinion (on their side).

Main probllem with the Court of Appeal is the waiting list. I had a case given permission this JUne and it has an indicative hearing date for November 2021. We still had to put the skeleton in four weeks after permission was granted though.

Any of that covid related or is it always that slow?

[Sheilbh]

Moving from annoyed to incandescant about the way they've worked out students' grades

A-Level results 2020: How have grades been calculated?
By Dave Thomson|13th August 2020|Exams and assessment

Updated 14th August at 10.40 to make an important clarification about adjustments for prior attainment not taking into account historic value added at centre level.

Schools up and down the country are trying to get their heads around the process that Ofqual, the exams regulator, has applied to calculate A-Level grades.

The process is described in some (but not full) detail in the 319 page Ofqual technical report and further detail can be gleaned from an accompanying data specification.

Awarding bodies have provided schools with standardisation reports, supposedly to show how they have arrived at the calculated grades. But these leave more questions than they answer in my view.

Here's an example from a school that has kindly shared a section of its report with us. To protect confidentiality the school and subject is unidentified.


In this subject, the left-hand table shows that 12.5% of entrants achieved A* between 2017 and 2019. And none achieved a grade U.

Now let's turn to the right-hand table. It implies that, based on the prior attainment of 2020 pupils and the historic relationship between prior attainment and A-Level results in this subject nationally and adjustments to achieve the desired national distribution of grades [1], the historic A* rate should be lowered by 6.79 percentage points. Or put another way, the calculated 2020 A* rate at the school is 12.50% – 6.79% = 5.71%. And although no one achieved U between 2015 and 2017, the model predicts that 2.30% should this year.

This immediately suggests that the prior attainment of the 2020 cohort was lower, on average, than that of the set of pupils from 2017 to 2019. However, there does not seem to be any way for schools to check this.

The prior attainment adjustment also does not appear to take account of historic value added at the school. So in schools with historically high value added, the prior attainment adjustment will result in grades being lowered.

Using the two tables above, we can calculate the 2020 expected grade distribution for this school in this subject.


Using column D, the allocation process begins. 27 pupils took the subject in 2020. Each pupil counts as 3.70%, as two A* grades (7.4%) would exceed the expected A* rate (5.71%). Consequently, just a single A* grade is awarded.

The full mix of grades awarded can be seen in the table below.


Unfortunately, one pupil is awarded a U. This seems rather harsh given that the model prediction is for fewer than one pupil (2.30%, when each pupil counts as 3.70%) to achieve this grade.

The table above headed "prior attainment adjustment" is absolutely fundamental to understanding how this year's grades have been calculated. Unfortunately, it raises more questions than it answers.

The technical documentation suggests that Ofqual uses an age-standardised version of GCSE average point score. This is a good idea as GCSEs have been reformed in recent years. Those entered in 2015 by 18 year olds who took A-Levels in 2017 were all graded A*-G. Those entered in 2018 by 18 year olds who took A-Levels in 2020 will mostly have been 9-1, with some A*-G grades.[2] Although there are point score conversions that are used conventionally, they are not strictly equivalent.

In order for schools to do detailed checking it would be helpful for them to be given the measures of prior attainment used and how they have been banded in each subject, both for 2020 pupils and for historic pupils. In addition, the national prediction matrices and details of adjustments made to pupils close to grade boundaries would be necessary.

Want to stay up-to-date with the latest research from FFT Education Datalab? Sign up to Datalab's mailing list to get notifications about new blogposts, or to receive the team's half-termly newsletter.
Notes

1. Ofqual runs some further analysis to identify pupils notionally close to grade boundaries to move some up a grade.

2. Anyone who tries to calculate value added using the 2019 Department for Education value added calculations is likely to find that it has fallen in 2020. This is because GCSE average point scores increased in 2019 as a result of more GCSEs being graded 9-1. Consequently, expected grades will be too high.

What's mattered to them is basically the "credibility" of the system over everything else - so the right amount of each grade have been handed out, just not necessarily to the right kids.

So basically it feels like if your school had no A* in 2017-19, even if you have an outstanding pupil they cannot get an A*. But even if your school had 0 Us in that period and you have a class full of kids who are doing well, one of them may get a U

I do a tutoring thing for kids from disadvantaged backgrounds and the student I tutored lost two full grades on their A levels due to the algorithm. They lost both of their university places - but were luckily able to find somewhere they're very happy with through clearing, but this is a disgrace.

[Josquius]

Yep. It's insane. It is divorced from all reality. They seem to be obsessed with not messing up the year by year stats despite the fact the reasons for this year being broken are obvious and will be recorded.

[Sheilbh]

Yep. It's insane. It is divorced from all reality. They seem to be obsessed with not messing up the year by year stats despite the fact the reasons for this year being broken are obvious and will be recorded.

Yeah. And I feel like worst case scenario of just accepting the teachers' grades (which are wildly inflated) is that employers (who care about A Levels? ) treat the 2020 results with an asterisk, but also probably realise the kids have come up in the worst possible year. The worst case scenarios of maintaining the "credibility of the system" is thousands of kids - especially disadvantaged kids - losing out on, you know, life opportunities.

[Sheilbh]

So the Tories have, inevitably, U-turned on this - if only there had been some way of knowing how this sort of policy would be seen by literally everyone

I think this means the only part of the UK that has managed to successfully moderate their exams is Wales which is, in large part, because they basically have the New Labour system (AS Levels etc) and kept coursework so there's a larger dataset for each pupil to fairly moderate. Of course no doubt we'll soon discover some glaring error that causes a u-turn but at the minute it's pretty striking.

Edit: Oh no Wales haven't done results yet and, shockingly, having watched England and Scotland spoil themselves have announced they won't use any algorithms

Edit: Also the DfE press office decided to release this photo of Gavin Williamson in his office with what appears to be a whip on his desk. Which is a choice

[Tamas]

 

[Tonitrus]

A memento from when he was Chief Whip I presume?

[Sheilbh]

A memento from when he was Chief Whip I presume?

Oh God - I'd not made that connection.

Christ, he's basic