Examination of Witnesses (Questions 940
- 959)
WEDNESDAY 29 APRIL 2009
Mr David Fillingham, Ms Rhona MacDonald, Mr Keith
Derbyshire and Dr Stephen Lorrimer
Q940 Lord Forsyth of Drumlean:
I understand that, of course, but if you then further adjust that
to deal with what you call "inequalities" does that
not actually have an equal and opposite effect? Perhaps not an
equal effect, but an opposite effect.
Mr Fillingham: I do not believe it does because
the inequalities element of the formula is looking to address
the fact that even though health services may be efficient, not
everybody actually accesses them in an equal way. In areas of
high deprivation it is often the case that people from the more
deprived parts of the population do not get access to healthcare
at all or access it less. Adding the health inequalities formula
recognises that the capitation formula which is based on historic
utilisation, may not be picking up all aspects and needs in the
population.
Q941 Lord Lang of Monkton:
I was going to ask about preventive medicine. Presumably in your
four-part formula you weight the population count higher than
everything else in expenditure on preventive medicine? Am I right
or wrong?
Ms MacDonald: No. One of the challenges—and
it relates to the last question—is that the formula until
recently has been heavily influenced by utilisation, so most of
the formula has related to data which is about the use of hospital
and community services. It is a more recent development that we
have actually looked at the impact of preventative services. I
understand the point about if your preventative services were
very good you might then have lower utilisation rates. Of course,
if your preventative services are very good, your needs would
be lower and the challenge is to actually be able to measure need.
The spending on preventative services in the NHS is actually relatively
small at the moment, but as part of the health needs adjustment
that is the kind of thing we are looking at, what proportion of
spend do you use the needs formula with, compared with the proportion
of spend which you have for the utilisation? I think that is one
of the challenges you will face if looking at formulae for Barnett.
It is about different spending streams and how you bring them
together.
Q942 Lord Lang of Monkton:
Could I just ask a supplementary? It really relates to Lord Forsyth's
question. It sounds as though you are going to penalise success
because the more efficient people are at using the funds given
to them, the less they will receive in subsequent years. Is that
the case?
Ms MacDonald: I think what would happen is that
over time I cannot imagine a situation where we have mastered
all health needs, so at the end of the day what you are trying
to do is to get the most efficient system which responds entirely
to needs. Because of the statistical techniques we use in modelling,
we take out things like the supply effect, but it is true to say
that everybody in the NHS is trying to reduce the use of hospitals
and have more preventative measures. The pace at which people
do that is different across the country, but the statistical modelling
techniques enable us to deliver the averages. That is a terrible
word to use and my colleagues might do better.
Mr Derbyshire: Can I come in there, because
I understand what is behind the question. It is a question which
goes back a long way in resource application formulae of a potential
perverse incentive. The original allocation formula in the seventies
was based on standardised mortality ratios in different parts
of the country, so people said, rather cleverly, that if you were
not very good at preventing death, if your death rate got higher,
you would get a higher share of the total pot. Theoretically,
that is a perverse incentive. It is very difficult to imagine
that a health service would actually deliberately worsen the outcomes
of the population to get more money. It just does not work like
that.
Q943 Lord Forsyth of Drumlean:
That is not the point. It is not that they would worsen their
performance but that the most efficient are penalised for improving
their performance?
Mr Derbyshire: They would get less money per
capita if their population was healthier, that is correct, and
if they are very good at improving the health of their population,
you could say that they would be penalised by having less money
in the future because they had a healthy population.
Ms MacDonald: Could I just add, I think also
it is about the relative spend between something like the use
of hospitals and preventative services. We spend far more on the
use of hospitals, so the formula is weighted much more to utilisation
than it is to preventative services. As people get better at improving
preventative services, then more of the weighting moves to that
formula. That does guard to some extent against the supposed perverse
incentive, so over time the weighting of individual elements of
the formula do not stay the same. There is not one formula, there
is a number of elements of the formula applied to different amounts
of money.
Q944 Lord Sewel:
Two questions, if I may. Firstly, in your paper you say: "It
is not currently possible on a technical basis to determine the
weighting for this health inequalities formula." Why? Ministers
decided on 15 per cent. Why? The second question is, has anyone
done any work along the lines that if the Scottish NHS was part
of your system how that would affect allocation to Scotland?
Mr Fillingham: Perhaps I could make a couple
of introductory comments on that and then I will ask my colleagues
to comment. As far as the health inequalities formula goes, which
is your first question, the research team last time they looked
at the formula spent quite a lot of time to see whether or not
we could incorporate a health inequality element within the utilisation
formula and they felt that technically it was not feasible. Therefore,
we developed this separate mechanism. We then looked at a range
of ways in which we could provide objective evidence to say what
the level of weighting should be, so we did consider, for example
in relation to an earlier question, the amount of money which
is currently spent on preventative services and should it relate
to that. We also looked at the differentials in ill-health between
different groups of the population. We concluded that this was
a matter of judgment and that the evidence base was not strong
enough to recommend to ministers a particular level, although
I think ACRA did consider a range of possibilities and 15 per
cent was within the range of possibilities we suggested would
be appropriate, but the process of actually deciding on 15 per
cent was subject to ministerial decisions.
Dr Lorrimer: The 15 per cent was a ministerial
decision and it was based on looking at the way the various options,
drawing on the ones which had been proposed by ACRA, distributed
funds across PCTs. It was a matter of judgment rather than anything
dramatically objective because there was this lack of evidence
as to what the weighting should be, so ultimately it had to be
a judgment for somebody and that fell to ministers. On the second
point about whether we looked at whether or not we should apply
this to Scotland, we have not. Firstly, we are an English department
and so that would not be part of our normal remit. It is also
not clear whether or not our formula would be appropriate for
Scotland.
Q945 Lord Sewel:
I appreciate that. I just wondered whether you were aware of any
sort of sad academic who has done it? We have seen lots of sad
academics who have done all sorts of various things.
Dr Lorrimer: I do not think I can think of any
sad, or indeed interesting academics who have looked at trying
to apply it for Scotland.
Mr Derbyshire: But they would if you paid them
sufficient money!
Q946 Lord Moser:
There are two parts of your paper which I find quite difficult
to understand as a statistician. I have to admit that. There are
two references where you shy away from trying to produce a formula
and both of those struck me. I think they are unrelated. The first
one relates to the first report combining age-related and additional
needs. What you are saying is that you tried to get a formula
which took account of the age composition of a population, not
just the total numbers, and that is a very relevant point to us,
but you shied away from it because for some reason you could not
do it and you ended up in fact having a different formula for
each age group, so you ended up with 18 formulae, if I understand
that rightly. Maybe that is inevitable. Maybe we have to end up
with 18 formulae. I hope not. The other one is the one which Lord
Sewel has already asked you about and I did not understand your
answer. When you were talking about health inequalities you said
that you really wanted to find a formula but for some reason,
which I think you have tried to explain but I did not understand
it, it was impossible and therefore judgment was brought into
the game and you ended up with 15 per cent. So two examples, which
perhaps we should take very seriously, where you ended up saying,
"No formula."
Mr Derbyshire: Yes. If I can try and understand
that in the context of Barnett, which is obviously about having
a formula which does a slightly different thing to what the NHS
formula needs to do, the first thing to actually make quite clear
is that we did not fail to take account of age in the formula,
and formulae used to allocate resources to the NHS prior to this
formula actually had four stages in the calculation of the weighted
capitation. The first was to count the population and then adjust
for age, because we know and have good quality data on the relationship
between age for the average individuals and their needs for health
services. Previous formulae had an adjustment for age and then
had an additional adjustment for need over and above the age effect.
What this formula does for the first time is to actually simultaneously
estimate the effects of age and need together, which is why we
consciously chose to have 18 different formulae for different
age bands because the additional needs drivers for each age band
are different. We have separate needs drivers for the 20-year-olds
and the 50-year-olds. So we did not fail to take account of age,
we just took account of age and need simultaneously.
Q947 Lord Moser:
I think I understand your answer, but the whole point for us is
that if one recognises something like the importance of age in
the revised Barnett Formula, say, you talk about statistical modelling
but it is not explained what you actually do. Is it not possible
to get all that together by weighting?
Mr Derbyshire: It is by simple weighting, and
previous health formulae in England and the health formulae used
in Scotland, Wales and Northern Ireland do first of all adjust
the population for age because demography plays a great part in
the relative needs for healthcare. The second order is adjusting
for relative need. All I am saying is that the current formula
in England, which is regarded as an advance, estimates the age
and need effects simultaneously.
Mr Fillingham: Perhaps, my Lord Chairman, I
could give an example to explain that. The research has shown
that there is a series of socioeconomic characteristics which
impact upon the demand for healthcare, but they differ in differing
age groups of the population. For example, if you look at young
people not staying on in education, that has an impact upon the
health needs of people in the 15—24 years age bracket. It
does not impact upon the health needs of people in the 65 plus
age bracket, whereas for pension credit claimants, for example,
the opposite would be true. So what the new formula does is to
look at the socioeconomic characteristics of the population in
those particular age bands. That is what Keith means by combining
age and need together.
Q948 Lord Rowe-Beddoe:
In your opening statement you suggested that your recommendations
from ACRA were, I think you said last time, accepted by the minister.
I am not quoting directly. Can you give examples of where your
recommendations in the last five years have not been accepted
and what was the result?
Mr Fillingham: Certainly since I became Chair
of ACRA, which is three years ago, our recommendations have been
accepted in full but there may be earlier instances.
Dr Lorrimer: In preparation for this Committee
we looked back through the records and we could not find any examples
where a recommendation had been rejected.
Q949 Lord Rowe-Beddoe:
I see. So your recommendations have been, as it were, taken up,
the prescription taken?
Mr Fillingham: Yes.
Q950 Chairman:
Except that there is a ministerial input, is there not? The 15
per cent, for example, was a figure the ministers produced, not
you?
Mr Fillingham: There is certainly a ministerial
input in a number of ways. The Secretary of State commissions
our work programme and also sets the objectives for the Committee,
so in the first instance there is a ministerial involvement there.
There are occasionally some issues which we say are beyond the
remit of ACRA or where we do not think there is the technical
evidence and the weighting on the health inequalities formula
was one of those, and of course the pace of change policy.
Q951 Lord Sewel:
It is critical, is it not? That is the big one. The pace of change
is the big one.
Mr Fillingham: Absolutely, because what we do
is determine the target allocation. The actual allocation which
is received is determined by the pace of change policy. I am conscious
we did not answer your earlier question about the health inequalities
formula. Just to be clear, there is a formula which determines
on the basis of health inequalities, using disability, free life
expectancy, what the differential is between different PCTs. For
that there is a formula. The issue there is how you put that together
with the needs formula to come to a bottom line. There is a number
of options there. You could multiply them or you could add them
together, and we recommended that they should be added together.
The issue of judgment was the scale, the emphasis which you placed
on health inequalities compared with the emphasis which you placed
on the utilisation element.
Q952 Chairman:
Could I just follow the 15 per cent for a second. I do not want
to leave the 15 per cent sort of up in the air. As I understand
it, you said to ministers, "Look, we can't really give you
a precise figure," but the ministers said, "Well, we
think 15 per cent is about right so that's what we're going to
do"?
Mr Fillingham: In actual fact—and I think
it is in our published report—we suggested, I think, four
options, which were not having a health inequalities formula at
all, 10 per cent, 15 per cent or 20 per cent, and there was a
series of rationales for each of those, but there was not strong
evidence to choose one or the other so we said we will not make
recommendations, to which ministers then went for the 15 per cent.
So it was not simply, "We can't make our minds up. What do
you think, Secretary of State?" There was a range of options,
but we did not recommend a single option.
Lord Forsyth of Drumlean: This is the
same point. How do you ensure the integrity and acceptability
of the system in those circumstances? I have to say that if I
were the minister I would be running the numbers and seeing which
marginal seats were affected, and things of that kind! Of course,
it might not influence my judgment!
Chairman: That is why you are not the
minister!
Q953 Lord Forsyth of Drumlean:
But I would be tempted and people might think that I would be
tempted. Surely the whole point of the system is that it should
be objective?
Mr Fillingham: Absolutely, that is the point
of the system and one of the key tests of ACRA is, is it felt
to be fair by the NHS, and I think we were therefore pleased with
this allocation round, but most of the commentary was suggesting
that the recommendations we had made which were accepted by the
Secretary of State were felt to be fair. Part of that is because
we can point to an evidence base for those recommendations and
people can see how that links back both to the need for healthcare
and to health inequalities.
Q954 Lord Forsyth of Drumlean:
But if you have a spread from nought to 20 per cent that is quite
a wide margin for adjustment?
Ms MacDonald: Could I add something? I think
in the papers one of the things we have made clear is that there
is no such thing as the perfect formula. We have been doing this
for a number of years and every year the aim is to be more refined
and get it better. One of the things which have been missing from
resource allocation has been any recognition—we have been
so dependent upon utilisation we have not been able to pick up
un-met needs, so nought would be an option which made no attempt
to recognise that actually there were factors of need which were
not addressed through utilisation of services. 20 per cent. What
that does is to give you an idea of how much of the NHS budget
would be focused on addressing inequalities, and that is not just
about preventative services, that is about whether in a hospital
people are treated equally, how you have to put special services
into an acute hospital to manage people's learning difficulties.
In giving the options to the Minister we were able to give him
some ideas about the issues which the inequalities adjustment
actually addresses, and we will do further work in order to improve
it. I think the question was very much about should we take a
step and how far should the step be? I think ACRA tried to produce
enough evidence to say that there is definitely merit in taking
a step towards inequalities. The service had long recognised that
the formula was not as good as it should be, so we had to go somewhere.
This is a journey and we have to keep pursuing it.
Q955 Lord Forsyth of Drumlean:
That is very helpful. It is just that in the paper it says that
in understanding from the ministerial decision how ministers decided
to target the 15 per cent, they targeted it to ensure the most
deprived areas have the resources they need, and that made me
think they were looking at tables which showed that it was not
ten, 15, 20 per cent, which PCTs would benefit. What you are saying
is that that was not the case?
Ms MacDonald: I do not know what the ministers
saw. What I am saying is that in presenting the information to
the ministers ACRA and TAG were able to look at the options we
were offering and do some empirical checking about how that looked
in relation to what we knew about deprivation. It was not a straightforward,
"These are the ones that gain and these are the ones that
don't gain." What you get are some interesting gainers and
losers when you apply the formula. Then what you have to do is
to look at that and say, "What does that say when we look
at what we know about services and what we know from other formulae?
What would that tell us? Does it feel fair?" So there is
an element of ACRA and TAG exercising a "felt fair"
judgment, but then recognising that actually we do not have the
evidence. It would not be peer reviewable, so we have to give
ministers a choice at that point. I have no idea and would not
want to comment on what went through their minds when making the
decision.
Q956 Lord Sewel:
I am just making mischief! May I just ask Dr Lorrimer, you, in
an aside, said you did not know whether the formula would be appropriate
for Scotland. If you devised a formula which allocates resources
according to criteria you set out, what on earth would be different
about Scotland that would make it not appropriate to apply it
there?
Dr Lorrimer: The first question would be, what
would be the objectives and the policy context for the Scottish
NHS as opposed to the NHS in England, and would those be different?
For instance, ACRA has these two objectives of meeting need and
also of addressing health inequalities. Those are set and commissioned
by the Secretary of State. In Scotland would those necessarily
be the same? I think you are right, that is probably a minor point.
I think probably more significant from a practical point of view
is that there will be two questions in my mind. One would be that
some of the issues in Scotland are different in the sense that,
for instance, the Scottish formula used at the moment includes
a rurality correction which ACRA could find no evidence we needed
in England, but it would be understandable why that might be more
important in Scotland. The other question would be actually are
the datasets which are available to us in England available in
Scotland or not, and vice versa, are the datasets which are used
in Scotland available in England? That would be the big practical
barrier.
Q957 Lord Rooker:
You said you used other formulae. With the local government system
in England there is a point in time when the decisions are announced
by ministers and there is actually an appeal window for local
authorities. Is there any similar system operating within the
Health Service?
Mr Fillingham: No, there is not a similar mechanism
linked to the allocations round of the NHS.
Q958 Lord Rooker:
So have you never been challenged by, let us say, a PCT which
felt especially aggrieved about something? Is there a means of
challenging it, and has that happened?
Mr Fillingham: Lots of people wish to draw their
views to ACRA's attention. They write to us, they commission research
papers, they seek to have various issues raised and presented,
and we have listened to those, but we do not take on board lobbying
because clearly our job is to try and use the best evidence which
is available and the best expert advice available to come up with
a formula that is as objective as possible. So there is not a
formal appeals mechanism. Yes, it is true that clearly PCTs will
have views on the process and their own position within the process
and they raise issues and on occasions we have commissioned research
to examine the issues they raise.
Q959 Lord Rooker:
I appreciate that. I was thinking more of when the decisions are
announced, if they are announced in principle. I do not quite
know how it is done, but at that point. Obviously during the ongoing
period of discussion people will always be bringing issues for
you to look at. One other point is that there is an awful lot
of money involved here in terms of the overall budget and the
formulae. I do not know whether it is probably that there is nothing
similar. Could I just ask you, you are all from the Health Department,
does the Treasury take an interest in what you do?
Mr Fillingham: First of all, Rhona and I are
from the National Health Service, not from the Department of Health.
It is a small distinction, but important to those in the NHS.
The Treasury reaches a settlement with the Department of Health.
ACRA's role then is to determine how the money which is allocated
to the NHS is allocated, so the Treasury does not have an active
involvement with ACRA.
| 
