Examination of Witnesses (Questions 209-219)
Professor Richard Tol
1 FEBRUARY 2005
Q209Chairman: Professor Tol, you are welcome.
I know you have been here sitting listening to how we proceed
and the way to handle this, but we are nevertheless extremely
grateful to you for coming along. As I said to your predecessors,
it is helpful to us if you speak up and relatively slowly so that
the record is as accurate as we can make it. Is there anything
you want to say in advance of questions or can we go straight
to questions?
Professor Tol: Only that it is an honour to
be here.
Q210Chairman: That is a very nice thing for
you to say, and we appreciate that. There are two aspects of your
work that particularly interest us: the work you have done on
the debate over the IPCC emissions scenarios; and the findings
of your Integrated Assessment Model "FUND" on costs
and benefits of tackling global warming. On the first, the emissions
scenarios, we know you are aware of the debate sparked by Professor
Henderson, who came and talked to us a little while ago, and Mr
Castles. What is your view on how this debate now stands?
Professor Tol: This is a very difficult issue
actually, so my answer will be a bit longer. The first point is
that all of the scenarios used in the climate change debate have
assumed convergence of per capita incomes. That was addressed
before, but that is perhaps not a valid assumption. But everybody
has assumed that so far. Therefore, we basically have to live
with that. It is not so that you can just take a scenario and
replace it with something else in a matter of five minutes; it
requires a lot of work. All these scenarios assume convergence.
Castles and Henderson are very right to criticise these scenarios
because they essentially assume convergence based on market exchange
rates, which is ludicrous I think. It should have been based on
purchasing power and exchange rates, and therefore these scenarios
are biased. It does not stop there. Castles and Henderson in their
critique unfortunately forgot that there is another convergence
assumption that has been made in all the scenarios, and that is
the convergence of energy efficiency, how much energy you use
in your economy. In all the scenarios that are there, convergence
is assumed as well. That is much more in line actually with the
historical evidence. The switch from MER to PPP for market exchange
rate to purchasing power exchange rates works the opposite direction
in energy efficiency because the gap in energy efficiency between
poor and rich is obviously much smaller, if you think that they
are much richer than you would think if you measured a market
exchange rate. So these two convergence estimates, and convergence
from market to purchasing power exchange rates actually offset
one another[2].
At the moment it is simply not clear whether they would fully
offset each other or whether they would hardly offset each other,
or somewhere in the middle. I think that Henderson and Castles
overstated that bias. The next issue is that the bias is there
and we do not really know how large it is, but we can compare
it to other uncertainties that are in the projections. I think
that most people working in this area would say that the error
or the bias introduced by (the exchange rate) or choice of exchange
rate is actually fairly small compared to the wider uncertainty
that there is about the future. It may be a small bias. That is
true on a global scale, if we are talking about global emissions.
It is not true on a regional scale because Castles and Henderson
are right that we should have projected slightly lower emissions
on a global scale, but the whole issue with exchange rates you
use does not really affect the rich countries but only the currently
poor countries. Therefore, global emissions may fall a little
bit, but the portion in those emissions that is due to developed
countries' emissions grows considerably and that of course changes
the whole political debate on who is responsible and so on and
so forth. The real issue is actually there. The final point—and
that is what I think is the real contribution of Castles and Henderson—that
they held the community that has developed these scenarios to
the light. They point out a great weakness in what has been done
in the IPCC but also in other corners of the climate and energy
community. That is where these scenarios come from, and who builds
them. If you look a little bit deeper into this, you find that
the SRES (and IPCC) scenarios were essentially built by people
who know a tremendous amount about energy supply. If you are interested
in energy supply the team that built the SRES scenarios are the
dream team, that is the people you want round the table if you
want to talk about energy supply. But if you then move to other
equally interesting issues such as energy demand or population
growth or economic growth, you would find that the SRES team was
perhaps lacking in expertise. Also, the modes that were used,
the SRES scenario exercises, were essentially models that have
been developed traditionally for energy policy analysis, and now
all of a sudden these models have been used to build scenarios,
and that is not what they were originally designed for. From that
perspective, historically, it was not a surprise that the SRES
scenarios would at some point be criticised by eminent economists
such as Henderson. It may have been perhaps a surprise that it
took so long! Another point came to light during this whole debate—and
I was at the very start of SRES, which was in 1996, and it was
one of the most controversial debates that I have ever seen. Going
back to the convergence assumption, from the very start onwards
it was clear that the SRES team had placed itself under constraints
of political correctness, that is to say because it is an IPCC
exercise it has to be reviewed by all the governments in the world,
and if you come up with scenarios in which the African countries,
which are a fairly large bloc (in the UN)—if they do not
grow fast enough, they will never approve our scenarios. So this
is what we cannot assume—and that self-censuring was right
there from the start.
Q211Chairman: You have given all those erudite
comments, and they sound very convincing to me, but I also form
a conclusion that whilst they may be right, at the end of the
day they do not make much difference either. The emissions we
are looking at are not so wildly out; they are just different,
but not sufficiently different. What are you saying?
Professor Tol: The switch from market exchange
to purchasing power exchange rate makes a relatively small difference
in global emissions. It makes a large difference in regional emissions.
But it particularly points us to a lot of work that will need
to be done in the future still. I was at a meeting discussing
this, three weeks ago in Washington DC. That is how you started
the question—where does the debate stand? After one and
a half days of vigorous debate we concluded there is something
fishy going on. We do not know whether it is a small fish or a
big fish.
Q212Lord Goodhart: One of the things we are
looking at very closely here is the question of costs and benefits
of climate change. Can you explain how damages from global warming
would be calculated?
Professor Tol: Yes, I can, because that is what
I have been doing for the last 15 years. Actually, it is very
simple—and that is a bit embarrassing because it is so simple.
One starts with estimating what we usually refer to as the physical
effects—how much faster would plants grow; how much higher
would the dykes need to be; how many people would be killed or
not get killed because of infectious diseases or heat stress,
or what-have-you? That is the first thing you estimate, things
that are physical or natural. The second thing one estimates is
what would be the price of these things, or the equivalent welfare
loss. It is easy if you are talking about dyke building and investment
costs for building the dykes, but if you are talking about loss
of human life it is somewhat trickier.
Q213Lord Goodhart: It must be very tricky. Do
you weight it according to the place and part of the world that
they are in? Do you value the cost of a life in western Europe
higher than you do in central Africa?
Professor Tol: That depends on the analysis
you are interested in. Originally these estimates are based on
local prices. That is perfectly acceptable to most of us, that
if you want to compare costs of dyke building in India or in Bangladesh
to costs of dyke building in Netherlands, you would use a local
price. By the exact same logic, if you are talking about loss
of life and you are using a local price, what would it mean to
them to have an increased risk of mortality or morbidity and what
would it mean to us? Initially these prices or values are local,
and then the next step is to find how you add them up and compare
them. That strongly depends on what decision-maker you are informing.
In a national context these things are well established. If you
are designing a policy on road safety in the United Kingdom, you
would never say that Scotsmen deserved less safety than Englishmen
because the Scots are poorer—or maybe they are richer now,
I do not know. You would never make such an argument because you
would get killed in Parliament. There is no need to explain that
here. But on a global scale you do not have that sort of mechanism,
so essentially you have to decide. If you were putting yourself
in a position of a global dictator who controls the whole world—a
benevolent dictator of course—or if you were to put yourself
in a position as somebody in a global parliament, then you would
argue that somebody dying in Bangladesh should be equally worthy
as somebody dying in the UK.
Q214Lord Goodhart: Even in this country, in
the UK, when somebody is killed in a road accident and their estate
sues for damages, they will get more damages if they were more
highly paid and lost more expectation of salary.
Professor Tol: Very true. That is what would
happen if you were talking about a market system or a court system,
but as a national government you would never design a policy that
way. It depends on the application and interest. From the perspective
of the global decision-maker, it is quite clear that you should
value everybody equally. From the perspective of a national decision-maker,
that is not true. What matters to the UK Government is how much
they value UK citizens relative to Bangladeshi citizens. That
is not necessarily related to how Bangladeshis value themselves.
Q215Lord Goodhart: Do you put any kind of evaluation
on things that produce an economic loss? For example, if one of
the consequences of global warming is the extinction of the polar
bear, do you put any value on that; and, if so, how do you do
it?
Professor Tol: There is a debate on that. In
my model I do it. A lot of people prefer not to do it. I do not
use values for the polar bear particularly. The basic method is
as follows. It is clear that people care about such species and
that people pay to go and see them and so on, so they represent
a value to human-beings—that is clear. The question is how
to estimate that. There are more reliable or less reliable methods
for doing so. If we are talking about species loss, unfortunately
we are forced to use the less reliable methods because the reliable
ones do not work; but, yes, they are included. Essentially, it
is based on questionnaires, asking people: "The polar bear
is threatened. Suppose that you could save it; how much would
you be willing to pay each year to help save the polar bear?"
Then you can calculate the risks and so on and so forth—what
would be the value of the polar bear. You can derive these values.
They are awfully soft, I immediately admit. Unfortunately, if
you do not do it, then implicitly you set the value to zero and
that is not what you want, so you use these methods, and because
the impacts of climate change on eco-systems and species is still
so vague, in my model I only use average values for these things.
Q216Lord Sheppard of Didgemere: I wonder if
you can help educate me on the concept of equity weights. Can
you explain how they are used and what the bottom-line effect
is on your model, and the justification?
Professor Tol: I already started to answer this
in the previous answer. Equity weights are essentially a trick
one uses to come from the regionally divergent prices to an aggregate.
The national or the regional effects are estimated in dollar terms
or whatever, and then you want to aggregate them, but then you
need to correct for the fact that the dollar to a poor man is
not the same as a dollar to a rich man. That is where the equity
weights come into play. You express things as an equivalent welfare
loss, and then add up and then convert back into dollars or pounds
or euros. That is essentially what equity weighing does. It reflects
the position of the global decision-maker, not the national or
regional decision-maker or individual, so in that sense they are
an academic construct. Equity weighing is terribly important,
though it is not clear how important it really is. It is important
because most people agree that developing countries and poorer
countries are more vulnerable to climate change than richer countries.
Therefore, if you use equity weighing you are placing more weight
on the damage in developing countries, and therefore the impact
estimates go up. The marginal damage cost estimates, so the additional
damage that is done by an additional tonne of carbon dioxide emitted,
go up by a factor of three or two of five. So the optimal carbon
tax would go up if you used equity weighting.
Q217Lord Skidelsky: I would like to go back
to an earlier part of this cost benefit analysis. Is there any
credible way of measuring emissions because the whole calculation
really depends on that? Everything else in the argument then depends
on it because the rate of global warming depends then partly on
the rate of emissions. If industry burns coal, you can infer that
a certain amount of CO2 will be released into the atmosphere,
but how much will obviously depend on the technical characteristics
of the coal-burners and so on. I can see that you know what the
world price of coal is, and you know how much coal is consumed,
and you can come to a dollar figure of total coal consumption
and therefore a figure for damage from CO2 emissions, assuming
a certain relationship between coal burning and CO2 emission.
However, how do you build in the technical characteristics of
the different types of production into this aggregated model?
I am just interested because it seems to me that unless you have
a credible method of measuring these things you also cannot begin
to start trading emissions if you want to go on to policy.
Professor Tol: There are two answers. First,
I am not a real expert in this, so maybe somebody else will correct
me. If we are talking about burning of fossil fuels it is fairly
easy. When you burn fossil fuels you oxidise the carbon that is
in the fuel, and that is the process of burning. You know how
much carbon is in a tonne of coal or a litre of oil or a cubic
metre of gas because that is basically why you burn it, so it
is fairly well known. If you are talking about cars, burning gasoline,
you either get carbon dioxide or carbon monoxide out, and you
know exactly how much of this, and you also know to a fairly close
degree what is the share of carbon monoxide and carbon dioxide.
That is fairly well under control. Then of course you can also
monitor how much carbon dioxide enters into the atmosphere, so
you can check. Things become much more complicated if you are
talking about emissions from agriculture, nitrous oxide and methane
and so on, and that is more sophisticated; but I am no expert
there. With regard to trading emission permits, what matters there
is not how much carbon is emitted physically but how much carbon
is emitted legally, because you are trading a legal document and
the regulator will have specified how to do this and how you account
for it. Whether that is anything to do with reality is a different
matter, but that is the legal reality.
Q218Lord Skidelsky: A poor country can say,
"we are legally entitled to do this; you say we are doing
that, and therefore we have to pay a penalty , but we deny it".
One can see all sorts of—
Professor Tol: Yes.
Q219Lord Skidelsky: I am sorry, I am taking
you beyond what we are talking about at the moment.
Professor Tol: You are right.
2 I should have said: If one switches from market
exchange rates to purchasing power parity, convergence of per
capita income as well as convergence of emission intensity change,
and these two changes offset one another. Back
|
