New AGFS proposals: The impact in context

16 January 2017

By Professor Martin Chalkley

Professor Martin Chalkley has advised the Bar Council on remuneration matters since 1995. He was one of the architects of the original Advocates' Graduated Fees Scheme, since which time he has been a consultant to the Bar Council on the AGFS, the CPS GFS, and family graduated fees. As a consultant member of the Bar's AGFS Working Group, Professor Chalkley has analysed and modelled all of the fee data.

Drawing on this experience and expertise, Professor Chalkley has written for the Bar Council to communicate critical details about how the new AGFS scheme has been developed and how its anticipated impact can usefully be evaluated.

The impact assessment accompanying the Ministry of Justice (MoJ) ' Reforming the Advocates' Graduated Fee Scheme' consultation paper contains quite a large number of figures. Some of these have already attracted commentary and discussion. One problem with interpreting the numbers is the lack of comparators; most of the numbers can only be understood when set in context. For example an "increase" of 10% might seem large, but when set against a previous reduction of 20% then maybe not so. The purpose of this note is to provide those who wish to respond to the consultation with some of the necessary context.

Graduated fees for criminal defence advocacy have been around for 20 years. There is therefore inevitably a lot of history and a lot of context. I cannot hope in this short note to give a comprehensive account of all that has happened and all that has changed leading up to the consultation, so I am going to be highly selective, but I hope the experience that I have had with the various schemes means that I can select wisely.

One issue is that since the MoJ has required any new scheme to be cost neutral, the proposals are all about redistributing a given expenditure. The architects of the scheme want that redistribution to be informed by certain principles - to make the fees more closely reflect the advocacy input required in different sorts of cases. So it is useful to start by thinking about how the present scheme has come about and what it does in terms of distributing fees.

Actually it is helpful to go right back to 1996 when the original Graduated Fee Scheme (GFS) was introduced. That scheme, designed by advocates for advocates, relied on the pre-existing fees mechanism (the ex post facto process) to inform how fees should be set relative to each other. The idea was that detailed and expert assessment of the work done in a large number of cases would provide a reliable guide to which cases required more or less work. That is pretty much how the original 1996 GFS was determined - fees reflect edevidence of the amount of skill and work required. You might therefore think in terms of the original GFS scheme fee table as providing some evidence of the relative value of different cases in 1996. No one, certainly not me, expected those relative values to stay fixed. Some cases have become more complex and onerous, others have ceased to exist, and new offences and cases have come along. I argued long and hard for a process of evidence-based review of the GFS but that never materialized. Instead fees were adjusted, revised and extended on the basis largely of financial and political imperative. Some fees that looked "high" were reduced relatively whilst others that looked "important" were increased, but there was really no evidence to support these changes. Yet the effects have been over the years very substantial and have dramatically changed the distribution of fees.

To give you a sense of this I will take you back to 2006 and Lord Carter's review. Fortunately, I still have the working figures from that review process. By 2006 the GFS had already changed a lot from 1996 but we have to start somewhere, so 2006 will have to do. I will start with the distribution of fees as between trials, cracks and guilty pleas. Going into Lord Carter's review, the proportionality between these - looking at overall spend and asking of the total spend on trials, cracks and pleas, what percentage went on each- was 58% for trials, 35% for Cracks and 6% for Guilty Pleas respectively. Now how did the "Revised" AGFS divide up the moneyfor those same cases?  The answer is 55% for trials, 34% for cracks and 11% for guilty pleas. That last figure is pretty remarkable. The revised scheme dramatically increased the relative fees of guilty pleas effectively doubling their share of the fee "cake". You might be tempted to ask why, if the previous expert assessment of bills suggested that guilty pleas were in aggregate worth 1/10 as much as trials, it was ever concluded that the figure should be 1/5. Why should there be such a huge relative fee rise for guilty pleas? That is a good question, and one I have asked myself and never received a satisfactory answer for.  

Whilst we are comparing how the cake is divided, here is another snippet from my spreadsheets from the Carter-era Revised Advocates Graduated Fee Scheme (RAGFS). Under the then pre-existing scheme, cases where. there was either a QC or two counsel took up a rather impressive 48% of the budget so that juniors alone were 52%. The intention with RAGFS was to reduce these two-counsel cases so the calculations for the new scheme indicated that those two counsel cases would decline to 36% of the budget. Juniors alone would get a substantially larger slice of the cake. But compare that to the figures in the consultation paper impact study. QCs, leading and led juniors currently receive 24% of the total and under the new proposals that would increase to 25%. This is a slightly bigger slice but still very much smaller than it was in 2007 and that in turn was a lot smaller than that in 2006.  So I hope this illustrates why context matters. On the one hand the figures in the impact study seem to indicate that guilty pleas will lose out, but the history is that they dramatically benefited from the 2007 revision and so by comparison the difference being proposed now is modest. Similarly whereas juniors alone look set to have a reduced slice, compared to the reductions that have been imposed over the years in payments to two-counsel cases, that looks like a small redistribution.

The Carter revised GFS was accompanied by increased funding. The story ever since has been one of fee reductions. Have those reductions been pro rata? You can already guess that the answer is no. The period since 2007 has been one of what might be described as ad hoc downward adjustment. Certain categories of case, certain roles of counsel, and different dispositions of cases have been subject to differential reductions.  To tell the full story would require more space than I have here but I can paraphrase a little. In essence from 2007 to 2013, longer more complex cases have been targeted for greater than average reductions in fees. Some figures help to give some sense of the magnitude and the variability of all these changes. I engaged with the Ministry of Justice in 2014 to try and understand exactly what the cumulative effects of cuts from 2007 up to that point had been. These figures are robust in the sense that they take a given bundle of cases (those cases paid for in 2013) and ask how much less they were paid than they would have been, had they been paid under the Carter Revised AGFS scheme. For trials the figure was 17%, so this means that from 2007 to 2013 trial fees were cut 17%. For guilty pleas the figure was 14% - bear in mind that guilty pleas were very favourably treated under the Carter reforms and continued to be relatively "protected".  For cracked trials you had better brace yourself - it was a 35% cut. This illustrates just what a lottery fee "adjustments" under GFS have become over time and gives you hopefully some context for the adjustments being suggested in the consultation. Incidentally the same sort of highly variable pattern of cuts applies if one looks at different offence groups. The same calculation applied to all Group A cases indicates that the cut from 2007 to 2013 was 25%, and that for Group D it was 11%.

The lottery in terms of adjustments to case fees that GFS has become is also a lottery for advocates' fee incomes. Here I think the impact assessment itself becomes problematic. To report that "on average" the impact of the new proposals on some large group of advocates will be x% distracts from the fact that within that group there will be many that face substantially larger or smaller changes based on their case-load. I have looked at the underlying data. Whereas for self-employed juniors alone the reported impact is -1% my figures indicate that about 60% of those juniors will gain under the new proposals and 40% will lose. For QCs for whom the reported impact is +10%, 70% will gain and 30% will lose. Of course some of those gains and losses will be larger than others giving us the reported average.

But bearing in mind what I have detailed above, what does that really tell us? The existing scheme is something of a lottery. The previous holders of the winning tickets will now face cuts; and vice versa.  Having criticised the notion of reporting on the average impact on different broad groups of advocates, I found myself wanting to report at least some figures by way of context. So for what it is worth (not a lot!) since 2007 the differential fee cuts that have been imposed reduce junior alone fees by 17% and QC fees by 22.4%.

From 1996 to 2006 there were much larger proportionate reductions in QC fees relative to juniors alone. Trying to make sense out of the relationship between the actual fee rates for QCs and juniors alone over the years is like trying to solve an ever changing Rubik's cube. The "ratio" between a QC fee and a junior alone fee has varied according to the type of case, the details of that case (the length of trial, PPE etc) and the disposition (trial crack and guilty), and it has varied a lot between different variants of GFS. To give some idea, QCs are nearly exclusively instructed in Category A cases. Take a 10 day murder trial. In 2005 a QC in such a case would have had a fee about 2.2 times higher than a junior alone. The comparison is I know spurious - the two would not have done the same case. Indeed that's the whole point of "Counsel Role" in that it is only when the case is serious and difficult enough that it gets two counsel or a QC. But, whatever - the ratio would have been 2.2. to 1. The 2007 scheme set the ratio for this case at 1.7 to one - so if you insist on seeing things this way that was a relative pay cut for QCs which, as the figures above suggest, is a process that continued.

Those tasked with reconsidering AGFS concluded that there was a need for a root and branch rethink about relative fee levels - that over 20 years the scheme had essentially lost any grounding in relative fees reflecting complexity of advocacy required. Hence the proposals are for a new approach and a new distribution, thought about from a "cases up" perspective posing the questions: What should the relative payment be for different types of cases?. Is that approach correct? Do the fees succeed in better capturing the requirements of advocacy? That is of course for advocates to decide and comment on.  For my part I would simply like to add two things. First, there is no sense in which the current AGFS can be thought of as reflecting an appropriate distribution of fees. It is best seen as an arbitrary and a distant distortion of any evidence-based approach. Second, this is rare opportunity for the profession to have its say and for that say to count. I hope this note in some way increases the information upon which that "say " is based.

Professor Martin Chalkley

If you have proposals that the Bar Council can consider before it finalises its own response, to any aspect of the scheme that you think could be improved, please send them to

As well as making your views known to the Bar Council, make sure that you communicate them directly to the Ministry of Justice in your own consultation response.