Examination of Witnesses (Questions 40
WEDNESDAY 23 JANUARY 2002
BENDER CB, MR
40. Am I right in saying that the company you
use, AEA Technology, say that the emission forecasts are subject
to uncertainties by up to 40 per cent?
(Mr Williams) Yes.
41. That is before you even put the figures
into the computer.
(Mr Williams) Yes.
42. So you start with a forecast of emissions
that are up to 40 per cent unreliable, which is not surprising
because if all of us could predict future economic growth we would
be making a killing on the stock market rather than being in Parliament.
However, you then take these things and put them into a computer
model to try to get a forecast of future air quality. That is
the next step, is it not?
(Mr Bender) Yes.
43. According to the NAO Report, once you have
put these figures in there are all sorts of things, mistakes in
computer programmes, misunderstandings of physical and chemical
processes involved, complexity of the physical and chemical processes
involved, the impact of causal variables, such as the weather.
Am I right?
(Mr Williams) Yes.
(Mr Bender) The pattern that you appear to be creating
is a static one. The implication of what you are saying is that
we do this as a one-off exercise over the next ten years and then
the world is fixed but that is not the case and if we were giving
that impression let me correct it. The inputs and elements of
the modelling are regularly updated. For example, in October we
looked at the issue of new emission factors for road transport.
We update the National Atmospheric Emissions Inventory on an annual
basis. We are dealing in an area of huge uncertainty in the way
you describe but it is not static, we are constantly reviewing
the modelling, reviewing the inputs to the modelling.
44. The weather is not static at all, the weather
changes every hour. The 40 per cent figure subject to uncertainty
is purely the emissions forecast. Am I correct in saying that
you do not even have an idea of what the uncertainties are with
the air quality forecast?
(Mr Williams) They will be broadly similar.
45. They might be more since you have put 40
per cent into an uncertain process.
(Mr Williams) Possibly more but there are reasons
for thinking they are not terribly bigger than that. Again, that
is inevitable in this sort of business. As Brian says, it is not
as if we forecast now and then sit back and wait ten years to
see what happens, we are constantly checking these models, we
are constantly monitoring. The monitoring data is the gold standard,
if you like, the measurements out there in the street. We are
constantly looking at that on an almost daily basis and checking
that against the models and so on. There is a constant refinement
and a constant check made. Again, it is back to the proportionality
argument. If we were spending half the GDP on some of these things
with that kind of uncertainty then you would ask why, and rightly
so. The point about the whole Strategy is to manage those uncertainties,
firstly to quantify them, and there is lots of forecasting in
environmental areas that do not quantify them at all, so at least
we know what they are and can act proportionately and accordingly.
The Committee suspended from 4.38 pm to 4.45
pm for a division in the House.
46. We were just on the uncertainty of the air
quality forecast and I wanted to pick up on something that Mr
Williams was saying. He said that there was evidence that the
uncertainties in the air quality forecasts were roughly the same
or perhaps a little bit bigger than the uncertainties in the emissions
forecast. In the NAO Report it does not give you anything, it
says AEA Technology has not tried to quantify a figure. Can you
explain briefly why you have your opinion that probably the uncertainty
is the same or a little bit larger?
(Mr Williams) The answer is rather technical so I
will not dwell on it. Essentially what it means is that what we
are forecasting are by and large annual averages and the change
on emissions to go into air quality is really the influence of
the weather and by and large the annual average weather does not
change hugely, certainly not as much as 40 per cent.
47. I really am a layman here. Do you not get
some years when you get very cold winters and very hot summers?
(Mr Williams) Yes, we do, and that is precisely the
sort of variation that we tried to bracket in the September consultation
last year, exactly that.
48. You accepted that this was a pretty uncertain
science and you said it would be difficult to justify if you were
spending half the GDP of the country on all of this, so I would
now like to ask you how much are we spending on this?
(Mr Williams) The figures are actually set out for
particles anyway in the September consultation document where
we analysed the effects of what we call an illustrative scenario
for control measures on transport emissions and on industrial
sources and so on. What we did there was to try and set the potential
costs of those measures against the benefits in terms of health
improvements expressed, where we could, in monetary terms.
49. What were the figures roughly?
(Mr Williams) I can read them out to you.
50. Just roughly.
(Mr Williams) The cost of the scenario was a range
there and it was of the order of several hundred thousand pounds.
Maybe I should just read them out for you.
51. While you are looking for those figures,
if I can turn to Mr Bender. On paragraph 3.23 it says that there
was an Interdepartmental Group on Costs and Benefits that did
look at two particular pollutants, nitrogen dioxide and particles,
and looked at options for reducing the levels. In London alone
they found that both options would have cost more than £100
million and even if they had been introduced would not meet the
air quality objectives. These are only two pollutants in one city,
albeit the largest city. You are talking about £100 million
and even then it would not achieve the targets that you have set.
(Mr Bender) But it would reduce towards that. Have
you got your data?
(Mr Williams) Yes. The illustrative scenario that
we talked about in terms of particles, the costs in annualised
terms for 2010 would have been between £785 million and £1,115
million. That is across the whole of the UK. What we then did
was try to calculate what an implied benefit value would be from
that bearing in mind the amount of life years saved and all that
sort of thing and set that against what we had estimated as the
value of a life year from other means and tried to set the two.
They broadly were commensurate.
52. I have to say I am a little bit sceptical
generally of regulatory impact assessments which try and make
these broad assessments. What you are saying is that this is just
(Mr Williams) Yes, that is right.
53. So for one pollutant the cost to the economy
is close to £1 billion or maybe more than £1 billion?
(Mr Williams) This was an illustrative scenario. That
is not something we were proposing or suggesting we do.
54. We are talking about at least nine pollutants,
maybe more depending on the European Directive, and we are talking
about targets that are arrived at through very uncertain science
where you have to factor in huge uncertainties such as the weather,
future economic growth and so on. Am I correct in saying that
there is almost no accurate cost benefit analysis of all the work
that your Department is doing on this?
(Mr Williams) Well, "accurate" is an uncertain
term in itself, I guess. What we have here are the best estimates
of the costs that we can come up with. We have come up with a
range of benefits. Admittedly there are uncertainties, yes. However,
the fact that we can get that far, I think, is no mean achievement.
There are lots of areas around the world that deal with these
sort of things that do not even get that far. Again it comes back
to this proportionality issue and where we can make the judgment
on the balance between costs and benefits, then we proceed along
a direction of proportionality. We are constantly trying to narrow
the range of those uncertainties and, for example, the Department
of Health, together with our own Department, is currently running
a project to improve our methods and techniques for monetizing
and valuing health impacts of pollution. We expect to see the
results of that sort of study later on this year. We are actively
seeking to narrow those ranges.
55. Of course, this is all being done to improve
human health and it says at the bottom of paragraph 2.14 about
a Health Department group: "However, it also concluded that
information on the health effects of exposure to nitrogen dioxide
and carbon monoxide was insufficiently reliable to be quantified."
So you do not even have any clear idea of the health effects that
all this policy and cost to the economy is imposing.
(Mr Bender) Again, the position on nitrogen dioxide
Mr Williams may want to add a bit to, but this is an area of difficult
science which is developing. We publish the basis on which we
produce these estimates. We do not produce them, as you recognise,
in a vacuum; it is produced across government with inputs from
Health, Treasury and other departments, and we try and approach
it on the basis of the best scientific advice we can get and the
best economic analysis we can get in what is an inherently uncertain
56. Am I right in sayingand I am being
a little facetiousthat the end result is a mini roundabout
in Manchester city centre?
(Mr Bender) If there is a particular local difficulty
then the solution in a particular area may include some local
traffic management, it may include a requirement for improvement
of vehicle emissions by road side tests and so on. It will be
a mix of things. It is conceivable the example you gave would
be true, but I suspect it would be rather more complicated than
57. I am not at all against improving air quality
but I am just concerned that this is a reach too far for the British
Civil Service. This is unreliable science leading to unreliable
targets based on unreliable evidence on the health effects. I
am concerned that a huge amount of time and money and effort by
good people is being wasted without any clear understanding that
this is improving the society we live in.
(Mr Bender) If I may say, I do think that is unfair.
Of course, this is a very complicated, very uncertain area. I
do believe that what we do in this country is in the forefront
of what happens internationally in these areas. The World Health
Organisation and the EU look at these things and it is not a trivial
issue dealing with 24,000 early deaths a year, so that is the
sort of issue we are trying to tackle.
Mr Osborne: Sadly my time is up.
Chairman: Mr Rendel?
58. Mr Osborne may not have known but I spent
three years in the atmospheric physics labs in Oxford studying
the ozone layer which at that time did not have a hole in it,
let alone anybody knowing about the hole. So at one time many
years ago I knew a little bit about this subject but probably
not now as most of it has long since gone, sadly. I wanted to
ask a bit about why you are studying the particular pollutants
you are studying and why you are not studying others. Perhaps
I can start with a question which is specifically to do with lead.
Lead can be a pollutant from a number of sources. To what extent
have you or your Department done any work on the possible double
effect of having lead from different sources in the human body
and whether that indicates you should be putting a greater emphasis
on lead in the atmosphere?
(Mr Williams) You are quite right, the Department
as a whole looks at the whole burden of lead in the environment
and how it affects humans through all the media, and there are
other parts of DEFRA that address that. What we were dealing with
here specifically was the air inhalation route and what we were
careful to do in incorporating lead into the Air Quality Strategy
is to have regard to those other routes of exposure through food
and water to make sure we were treating the thing proportionately.
59. When you say you "had regard"
to that, what did that mean in terms of how you tried to carry
out your policy?
(Mr Williams) What it meant was, put very simply,
that we did not require unnecessarily large or overly stringent
air quality objectives when there could be much larger routes
of exposure through other media. In a sense, the problem from
air, of course, is solved itself because the airborne emissions
of lead are now very, very small, virtually non-existent.