International environmental problems

1
chapter 9

• International environmental problems

2
Learning objectives

• How do international environmental problems
  differ from national (or sub-national) problems?
• What additional issues are raised by virtue of an
  environmental problem being international?
• What insights does game theory bring to our
  understanding of international environmental
  policy?
• What determines the degree to which cooperation
  takes place between countries and policy is
  coordinated? Put another way, which conditions
  favour (or discourage) the likelihood and extent
  of cooperation between countries?
• Why is cooperation typically a gradual, dynamic
  process, with agreements often being embodied in
  treaties or conventions that are general
  frameworks of agreed principles, but in which
  subsequent negotiation processes determine the
  extent to which cooperation is taken?
• Is it possible to use such conditions to explain
  how far efficient cooperation has gone concerning
  upper-atmosphere ozone depletion, and global
  climate change?

3
Game theory analysis

• Game theory is used to analyse choices where the
  outcome of a decision by one player depends on
  the decisions of the other players, and where
  decisions of others are not known in advance.
• This interdependence is evident in environmental
  problems.
• For example, where pollution spills over national
  boundaries, expenditures by any one country on
  pollution abatement will give benefits not only
  to the abating country but to others as well.
• Similarly, if a country chooses to spend nothing
  on pollution control, it can obtain benefits if
  others do so.
• So in general the pay-off to doing pollution
  control (or not doing it) depends not only on
  ones own choice, but also on the choices of
  others.
• We use game theory to investigate behaviour in
  the presence of global or regional public goods.
• The arguments also apply to externalities that
  spill over national boundaries.

4
Two-player binary-choice games
5
Ys strategy Xs strategy Strategy 1 Strategy 2
Strategy 1 a, a b, c
Strategy 2 c, b d, d
Figure 9.1 Two player binary choice games
6
Ys strategy Xs strategy Pollute Abate
Pollute 0, 0 5, -2
Abate -2, 5 3, 3
Figure 9.2 A two-player pollution abatement game
The payoff matrix here has a structure of payoffs
known as a Prisoners Dilemma X and Y are two
countries, each of which faces a choice of
whether to abate pollution or not to abate
pollution (labelled Pollute). Pollution
abatement is assumed to be a public good so that
abatement by either country benefits both.
Abatement comes at a cost of 7 to the abater, but
confers benefits of 5 to both countries. If both
abate both experience benefits of 10 (and each
experiences a cost of 7).
7
Characteristics of this solution

• The fact that neither country chooses to abate
  pollution implies that the state of the
  environment will be worse than it could be.
• The solution is also a Nash equilibrium.
• A set of strategic choices is a Nash equilibrium
  if each player is doing the best possible given
  what the other is doing.
• Put another way, neither country would benefit by
  deviating unilaterally from the outcome, and so
  would not unilaterally alter its strategy given
  the opportunity to do so.
• The outcome is inefficient. Both countries could
  do better if they had chosen to abate (in which
  case the pay-off to each would be three rather
  than zero).

8
Why has this state of affairs come about?

• The game has been played non-cooperatively. We
  shall examine shortly how things might be
  different with cooperative behaviour.
• The second concerns the pay-offs used in Figure
  9.2. These pay-offs determine the structure of
  incentives facing the countries. They reflect the
  assumptions we made earlier about the costs and
  benefits of pollution abatement. In this case,
  the incentives are not conducive to the choice of
  abatement.
• The pay-off matrix in Figure 9.2 is an example of
  a so-called Prisoners Dilemma game. The
  Prisoners Dilemma is the name given to all games
  in which the pay-offs, when put in ordinal form,
  are as shown in Figure 9.3.

9
Ys strategy Xs strategy Pollute Abate
Pollute 2, 2 4, 1
Abate 1, 4 3, 3
Figure 9.3 The two-player pollution abatement
Prisoners Dilemma game ordinal form
10

• In all Prisoners Dilemma games, there is a
  single Nash equilibrium.
• This Nash equilibrium is also the dominant
  strategy for each player.
• The pay-offs to both countries in the dominant
  strategy Nash equilibrium are less good than
  those which would result from choosing their
  alternative, dominated strategy.
• As we shall see in a moment, not all games have
  this structure of pay-offs.
• However, so many environmental problems appear to
  be examples of Prisoners Dilemma games that
  environmental problems are routinely described as
  Prisoners Dilemmas.

11
A cooperative solution

• Suppose that countries were to cooperate, perhaps
  by negotiating an agreement.
• Would this alter the outcome of the game?
• Intuition would probably lead us to answer yes.
  If both countries agreed to abate and did what
  they agreed to do pay-offs to each would be 3
  rather than 0.
• In a Prisoners Dilemma cooperation offers the
  prospect of greater rewards for both countries,
  and in this instance superior environmental
  quality. 
• But this tentative conclusion is not robust.

12
Sustaining the cooperative solution

• Can these greater rewards be sustained?
• If self-interest governs behaviour, they probably
  cannot.
• To see why, note that the Abate, Abate outcome
  is not a Nash equilibrium.
• There is no external authority with the authority
  to impose a binding agreement.
• Moreover, this agreement is not
  self-enforcing. Each country has an incentive
  to defect from the agreement to unilaterally
  alter its strategy once the agreement has been
  reached.

13
Other forms of game

• Not all games have the structure of the
  Prisoners Dilemma (PD).
• Even where a game does have a PD pay-off matrix
  structure, the game may be played repeatedly.
• As we shall see later, repetition substantially
  increases the likelihood of cooperative outcomes
  being obtained.
• Furthermore, there may be ways in which a PD game
  could be successfully transformed to a type that
  is conducive to cooperation.
• We now look at some other game structures.

14
Ys strategy Xs strategy Pollute Abate
Pollute -4, -4 5, -2
Abate -2, 5 3, 3
Figure 9.4 A two-player Chicken game
15
Figure 9.5 Extensive form of Chicken game
Pollute

Ys choice
(-4, -4)
Abate
Pollute
(5, -2)
Xs choice
(-2, 5)
Pollute
Abate
Abate
(3,3)
16
Leadership

• A strategy in which both countries abate
  pollution could be described as the collectively
  best solution to the Chicken game as specified
  in Figure 9.4 it maximises the sum of the two
  countries pay-offs.
• But that solution is not stable, because it is
  not a Nash equilibrium. Given the position in
  which both countries abate, each has an incentive
  to defect (provided the other does not).
• A self-enforcing agreement in which the structure
  of incentives leads countries to negotiate an
  agreement in which they will all abate and in
  which all will wish to stay in that position once
  it is reached does not exist here.
• However, where the structure of pay-offs has the
  form of a Chicken game, we expect that some
  protective action will take place. Who will do
  it, and who will free-ride, depends on particular
  circumstances.
• Leadership by one nation (as by the USA in the
  case of CFC emissions reductions) may be one
  vehicle through which this may happen.
•  

17
Bs strategy As strategy Do not Contribute Contribute
Do not contribute 0, 0 0, -8
Contribute -8, 0 4, 4
Figure 9.6 A two-player Assurance game
18
Games with multiple players
19
Our example

• Most international environmental problems involve
  several countries and global problems a large
  number.
• Much of what we have found so far generalises
  readily to problems involving more than two
  countries.
• Let N be the number of countries affected by some
  environmental problem, where N 2. For
  simplicity, we assume that each of the N
  countries is identical.
• We revisit the Prisoners Dilemma example.
• As before, each unit of pollution abatement comes
  at a cost of 7 to the abating country it confers
  benefits of 5 to the abating country and to all
  other countries.
• For the case where N 10, the pay-off matrix can
  be described in the form of Table 9.1.

20
(No Transcript)
21
Structure of pay-offs

• The structure of pay-offs is critical in
  determining whether cooperation can be sustained.
  
• Following Barrett (1997), we explore the pay-offs
  to choices in a more general way.
• Denote NBA as the net benefit to a country if it
  abates and NBP as the net benefit to a country if
  it pollutes (does not abate).
• Let there be N identical countries, of which K
  choose to abate.
• We define the following pay-off generating
  functions
• NBP a bK
• NBA c dK
• where a, b, c and d are parameters.

22
Structure of pay-offs

• NBP a bK
• NBA c dK
• By altering these parameter values, we generate
  different pay-off matrices.
• For the problem in Table 9.1 we have a 0, b
  5, c 7 and d 5.

23
Figure 9.7.
Figure 10.7 The payoffs to one country from
abating and from not abating as the number of
other countries abating varies.

24
Figure 9.8 The payoffs to one country from
abating and from not abating as the number of
other countries abating varies alternative set
of parameter values

25
Figure 9.9 The payoffs to one country from
abating and from not abating as the number of
other countries abating varies third set of
parameter values (a 0, b 5, c 3 and d 3)

26
Continuous choices about the extent of abatement

• We can generalise the discussion by allowing
  countries to choose or rather negotiate
  abatement levels.
• This can be done with some simple algebra. We
  leave you to read this for yourself here just
  report some key results.
• Non-cooperative behaviour
• Each country chooses its level of abatement to
  maximise its pay-off, independently of and
  without regard to the consequences for other
  countries. That is, each country chooses its
  abatement level, z, conditional on z being fixed
  in all other countries.
• The solution each country abates up to the point
  where its own marginal benefit of abatement is
  equal to its marginal cost of abatement.

27
Full cooperative behaviour

•  
• Full cooperative behaviour consists of N
  countries jointly choosing levels of abatement to
  maximise their collective pay-off. This is
  equivalent to what would happen if the N
  countries were unified as a single country that
  behaved rationally.
• The solution abatement in each country is chosen
  jointly to maximise the collective pay-off.
• This is the usual condition for efficient
  provision of a public good. That is, in each
  country, the marginal abatement cost should be
  equal to the sum of marginal benefits over all
  recipients of the public good.
• The full cooperative solution can be described as
  collectively rational it is welfare-maximising
  for all N countries treated as a single entity.
  If a supranational government existed, acting to
  maximise total net benefits, and had sufficient
  authority to impose its decision, then the
  outcome would be the full cooperative solution.

28
Figure 9.10 A comparison of the non-cooperative
and full cooperative solutions to an
environmental public good problem

MB
MCi
MBi
Z
ZN
ZC
29
International environmental agreements (IEAs)

• Role of UN framework linkage of issues -
  environmental protection, environmental
  sustainability and economic development.
• But much of what is important has been dealt with
  at regional or bilateral levels, and takes place
  in relatively loose, informal ways.
• The need for international treaties arises from
  the fact that political sovereignty resides
  principally in nation states.
• The European Union may be a challenge to that
  proposition.
• In the absence of a formal supranational
  political apparatus with decision-making
  sovereignty, the coordination of behaviour across
  countries seeking environmental improvements must
  take place through other forms of international
  cooperation such as formal international
  treaties.

30
Effectiveness of IEAs

• Judging effectiveness of IEAs requires
  construction of a counter-factual what would
  have happened anyway if the IEA had not been
  reached.
• Then, abatements achieved under an IEA can be
  compared with the counter-factual estimated
  abatement levels in the absence of the treaty.
• In other words, the test of effectiveness of
  agreements is by comparison of the Nash (non
  cooperative) and cooperative outcomes.
• By this criterion, the literature on IEAs
  suggests that they are likely to be very limited
  in their effectiveness.

31
Key results

• Three assertions about the effectiveness of IEAs
  seem to emerge from the theoretical literature,
  all of which imply somewhat pessimistic results
• Treaties tend to codify actions that nations were
  already taking. Or, put another way, they largely
  reflect what countries would have done anyway,
  and so offer little net improvement.
• When the number of affected countries is very
  large, treaties can achieve very little, no
  matter how many signatories there are.
• Cooperation can be hardest to obtain when it is
  needed the most.

32
Box 9.1 Conditions conducive to effective
cooperation between nations in dealing with
international environmental problems

• The existence of an international political
  institution with the authority and power to
  construct, administer and (if possible) enforce a
  collective agreement.
• The output of the international agreement would
  yield private rather than public goods.
• A large proportion of nation-specific or
  localised benefits relative to transnational
  benefits coming from the actions of participating
  countries.
• A small number of cooperating countries.
• Relatively high cultural similarity among the
  affected or negotiating parties.
• A substantial concentration of interests among
  the adversely affected parties.

• The adoption of a leadership role by one
  important nation.
• Low uncertainty about the costs and benefits
  associated with resolving the problem.
• The agreement is self-enforcing.
• Continuous relationship between the parties.
• The existence of linked benefits.
• The short-run cost of implementation is low, and
  so current sacrifice is small.
• High proportion of the available benefits are
  obtained currently and in the near future.
• Costs of bargaining small relative to gains

33
Other factors conducive to international
environmental cooperation

• Role of commitment
• an unconditional undertaking made by an agent
  about how it will act in the future, irrespective
  of what others do.  
• credibility of commitments
• one interesting form of commitment is the use of
  performance bonds
• obtain the benefits of free-riding on the others
  pollution abatement.
• Transfers and side-payments
• e.g. signatories offer side-payments to induce
  non-signatories to enter
• Linkage benefits and costs and reciprocity
•  may be possible to secure greater cooperation if
  other benefits are brought into consideration
  jointly. Doing this alters the pay-off matrix to
  the game.
• e.g. international trade restrictions,
  anti-terrorism measures, health and safety
  standards may be economies of scope available by
  linking these various goals.  

34
Figure 9.11 A one shot Prisoners Dilemma game
Bs strategy As strategy Defect Cooperate
Defect P, P T, S
Cooperate S, T R,R
35
Figure 9.12 The two-shot Prisoners Dilemma game
Bs strategy As strategy Defect Cooperate
Defect 2P, 2P TP, SP
Cooperate SP, TP RP,RP
36
Are players only concerned with the returns that
they get?
37
Global climate changeWhat determines Earths
climate?
38
Figure 9.13 Estimate of the Earths annual and
global mean energy balance
Over the long term, the amount of incoming solar
radiation absorbed by the Earth and atmosphere is
balanced by the Earth and atmosphere releasing
the same amount of outgoing longwave radiation.
About half of the incoming solar radiation is
absorbed by the Earths surface. This energy is
transferred to the atmosphere by warming the air
in contact with the surface (thermals), by
evapotranspiration and by longwave radiation that
is absorbed by clouds and greenhouse gases. The
atmosphere in turn radiates longwave energy back
to Earth as well as out to space. Source FAQ
1.1, Figure 1. http//www.ipcc.ch/pdf/assessment-r
eport/ar4/wg1/ar4-wg1-faqs.pdf.
39

• How would GHG emissions and atmospheric
  concentrations change over the coming century and
  beyond if no additional controls were imposed?

40
Fig 9.14 Global GHG emissions (in GtCO2-eq per
year) in the absence of additional climate
policies.
The figure shows six illustrative SRES marker
scenarios (coloured lines) and 80th percentile
range of recent scenarios published since SRES
(post-SRES) (gray shaded area). Dashed lines show
the full range of post- SRES scenarios. The
emissions include CO2, CH4, N2O and
F-gases. Source Figure 3.1, IPCC AR4 Synthesis
Report, available online at http//www.ipcc.ch/p
df/assessment-report/ar4/syr/ar4_syr.pdf
41

• How will climate change over the coming century
  and beyond?
•  

42
Figure 9.15 Multi model averages and assessed
ranges for surface warming
43
Options available for mitigating GHG atmospheric
concentrations

• There are two ways to move towards a goal of
  reducing the rate of growth of atmospheric
  greenhouse-gas concentrations
• increase the capacity of sinks that sequester
  carbon dioxide and other greenhouse gases from
  the atmosphere
• decrease emissions of greenhouse gases below
  business as usual (thereby reducing GHG inflows
  to the atmosphere).

44
The costs of attaining GHG emissions or
atmospheric concentration targets key results

• The cost of achieving any given target in terms
  of levels of allowable GHG emissions or
  stabilised GHG concentrations increases as the
  magnitude of the emissions or concentration
  target declines.
• Other things being equal, the cost of achieving
  any given target increases the higher are
  baseline (i.e. uncontrolled) emissions over the
  time period in question.
• The cost of achieving any given target varies
  with the date at which targets are to be met, but
  does so in quite complex ways. It is not possible
  to say in general whether fast or early control
  measures are more cost-effective than slow or
  late controls.

45
More key results

• There is some scope for GHG emissions to be
  reduced at zero or negative net social cost. The
  magnitude of this is uncertain. It depends
  primarily on the size of three kinds of
  opportunities and the extent to which the
  barriers limiting their exploitation can be
  overcome
• overcoming market imperfections (and so reducing
  avoidable inefficiencies)
• ancillary or joint benefits of GHG abatement
  (such as reductions in traffic congestion)
• double dividend effects

46
More key results (2)

• Abatement costs will be lower the more
  cost-efficiently that abatement is obtained. This
  implies several things
• Costs will be lower for strategies that focus on
  all GHGs, rather than just CO2, and are able to
  find cost-minimising abatement mixes among the
  set of GHGs. It is not just carbon emissions or
  concentrations that matter.
• Costs will be lower for strategies that focus on
  all sectors, rather than just one sector or a
  small number of sectors. Thus, for example, while
  reducing emissions in energy production is of
  great importance, the equi-marginal principle
  suggests that cost minimisation would require a
  balanced multi-sectoral approach.
• The more complete is the abatement effort in
  terms of countries involved, the lower will be
  overall control costs. This is just another
  implication of the equi-marginal cost principle,
  and it also is necessary to minimise problems of
  carbon (or other GHG) leakage.
• The above comments imply that in principle
  achieving targets at least cost could be brought
  about by the use of a set of uniform global GHG
  taxes. Alternatively, use could be made of a set
  of freely tradeable net emissions licenses (one
  set for each gas, with tradability between sets
  at appropriate conversion rates), with quantities
  of licenses being fixed at the desired
  cost-minimising target levels.
• Climate-change decision-making is essentially a
  sequential process under uncertainty. The value
  of new information is likely to be very high, and
  so there are important quasi-option values that
  should be considered.

47
Figure 9.16 Global GHG emissions for 2000 and
projected baseline emissions for 2030 and 2100
from IPCC SRES and the post-SRES literature The
figure provides the emissions from the six
illustrative SRES scenarios. It also provides the
frequency distribution of the emissions in the
post-SRES scenarios (5th, 25th, median, 75th,
95th percentile), as covered in Chapter 3.
F-gases cover HFCs, PFCs and SF6.
48
Numerical estimates of mitigation potential and
mitigation costs Short to medium term GHG
mitigation estimated mitigation costs for the
period to 2030  
49
(No Transcript)
50
Long term GHG mitigation, for stabilised GHG
concentrations estimated mitigation costs for
the period after 2050
51
Figure 9.17
52
Fig 9.17 (alt version)
53
Figure 9.17 Emissions pathways of mitigation
scenarios for alternative groups of stabilization
targets Accompanying text The pink area gives
the projected CO2 emissions for the recent
mitigation scenarios developed post-TAR. Green
shaded areas depict the range of more than 80 TAR
stabilization scenarios (Morita et al., 2001).
Category I and II scenarios explore stabilization
targets below the lowest target of the
TAR. Source IPCC (2007), WGIII. (Based on
Nakicenovic et al., 2006, and Hanaoka et al.,
2006)
54
Nordhaus DICE-2007 model

• An intertemporal optimisation model of climate
  change policy.
• Objective function in DICE-2007 is the present
  value of global consumption.
• Damages from GHG emissions reduce consumption
  possibilities, as do the costs of GHG abatement.
• Model allows the user to identify emissions
  abatement choices (the policy instruments) that
  maximise the present value of global consumption,
  net of GHG damages and abatement costs, over
  horizons of up to 200 years or so ahead. Nordhaus
  calls such a set of policy choices the optimal
  policy.

55
Table 9.11 Results of DICE-2007 simulations
56
(2005 U.S. dollars per ton of carbon)
Figure 9.18 The level of carbon prices (or taxes)
through time for various mitigation strategies
57
Safe minimum standard (precautionary) approaches.

• The large uncertainties which exist in climate
  change modelling regarding the damages that
  climate change could bring about lead many to
  conclude that mitigation policy should be based
  on a precautionary principle.
• This would entail that some safe threshold
  level of allowable climate change is imposed as a
  constraint on admissible policy choices.
• Support for a safe minimum standard approach in
  the climate change context has grown in recent
  years for two main reasons.
• First, the science increasingly points to
  non-linearities in the dose-response function
  linking temperature change to induced damages,
  with damages rising at increasingly large
  marginal rates at higher levels of global mean
  temperatures, and possibly discontinuously.
• Secondly, positive feedbacks in the linkage
  between GHG concentration rates and temperature
  responses are increasingly likely to kick-in as
  atmospheric GHG concentrations rise, so that the
  climate sensitivity coefficients rise
  endogenously.

58
Figure 9.19
59
The Kyoto Protocol

• Attempts to secure internationally coordinated
  reductions in GHG emissions have taken place
  largely through a series of international
  conventions organised under the auspices of the
  United Nations.
• 1992 Earth Summit Framework Convention on
  Climate Change (FCCC) was adopted, requiring
  signatories to conduct national inventories of
  GHG emissions and to submit action plans for
  controlling emissions.
• By 1995, parties to the FCCC had established two
  significant principles emissions reductions
  would initially only be required of
  industrialised countries second, those countries
  would need to reduce emissions to below 1990
  levels.

60
The Kyoto Protocol (2)

• Kyoto Protocol the first substantial agreement
  to set country-specific GHG emissions limits and
  a timetable for their attainment.
• To come into force and be binding on all
  signatories, the Protocol would need to be
  ratified by at least 55 countries, responsible
  for at least 55 of 1990 CO2 emissions of FCCC
  Annex 1 nations.
• The key objective set by the Protocol was to cut
  combined emissions of five principal GHGs from
  industrialised countries by 5 relative to 1990
  levels by the period 20082012.
• The Protocol did not set any binding commitments
  on developing countries.

61
Subsequent activity

• Since 1997, there have been annual meetings of
  the parties that signed the Kyoto Protocol.
• Initially, those meetings were largely concerned
  with the institutional structures and mechanisms
  and rules of the game required to implement the
  protocol, such as how emissions and reductions
  are to be measured, the extent to which CO2
  absorbed by sinks will be counted towards Kyoto
  targets, and compliance mechanisms.
•  The twin conditions required for the Protocol to
  become operational were met in early 2005. While
  the Kyoto Protocol came into force at that time,
  it did so without the participation of the USA,
  thereby significantly weakening its potential
  impact.
• The first phase of the Kyoto Protocol will end in
  2012.
• Recent meetings of the parties have been
  concerned with making preparations for its second
  phase.

62
The Kyoto Protocols flexibility mechanisms

• These generate incentives for control to take
  place in sources that have the lowest abatement
  costs, and so create the potential for greatly
  reducing the total cost of attaining any given
  overall policy target.
• Emissions Trading Allows emissions trading
  among Annex 1 countries countries in which
  emissions are below their allowed targets may
  sell credits to other nations, which can add
  these to their allowed targets.
• Banking Emissions targets do not have to be met
  every year, only on average over the period
  20082012. Moreover, emissions reductions above
  Kyoto targets attained in the years 20082012 can
  be banked for credit in the following control
  period.
• Joint Implementation JI allows for bilateral
  bargains among Annex 1 countries, whereby one
  country can obtain Emissions Reduction Units
  for undertaking in another country projects that
  reduce net emissions, provided that the reduction
  is additional to what would have taken place
  anyway.
• Clean Development Mechanism By funding projects
  that reduce emissions in developing countries,
  Annex 1 countries can gain emissions credits to
  offset against their abatement obligations.
  Effectively, the CDM generalises the JI provision
  to a global basis. The CDM applies to
  sequestration schemes (such as forestry
  programmes) as well as emissions reductions.

63
The Kyoto Protocols flexibility mechanisms

• Problems validation of project additionality.
• Kyotos flexibility mechanisms appear to offer
  very large prospects of reductions in overall
  emissions abatement costs. Studies carried out
  during the 1990s found median marginal abatement
  costs in developed economies to be of the order
  of 200 per tonne of carbon.
• Barrett (1998) argued that with emissions being
  uncontrolled in the non-Annex 1 countries,
  marginal abatement costs there are effectively
  zero.
• On this basis, he suggests that cost savings from
  the Clean Development Mechanism alone could be of
  the order of 200/tC at the margin.