When the cold war was at its height, the concept of security revolved around deterrence and the two super powers. One wanted to domesticise the universe, the other to universalize the domestic working class militancy. In the middle of the eighties, a new awareness emerged to questioning the relevance of mindless militancy built up as a means to define national security on a maximum – minimum continuum paradigm. In the US the discontentment of public brewed oven the military expenditure that could be expended on maximizing their domestic imperatives in terms of inordinate consumption, exorbitant lifestyle and production of wealth – in Soviet Union, questioning the logic of military expenditure without providing the minimum needs to its people came to the fore with winding down of its iron curtain.

In case of Environment, the concept of Securitization can be harked back to 1960s, when a number of popular books such as Rachel Carson‟s „Silent Spring‟ caused rethinking of the entire established relations between nature and civilization.

Further, host of books by writers like Paul, Enlirch, Garret Handin, Garry Commoner, Lester Brown erupted a public debate about the issues such as exponential population growth, the tragedy of commons, negative externalities of the production technology and complex global interdependencies of late twentieth century. The creation of World Commission on Environment Development in 1983, chaired by Norway‟s Brundtland and the release of its report „Our Common Future‟ was the result of this.

Environment-conflict linkage expostulated by Homes Dixon and others with an array of case studies share as in Rwanda, Somalia, Pakistan, Philippines unearthed the environment causes of conflict.

According to their studies environmental scarcity of resources has produced many social effects in terms of economic decline, poverty, decrease in agricultural productivity, migration and weakening of state structure which in turn result in violent intra, or inter state conflicts such as scarcity, group identify conflict and insurgencies and relative deprivation conflict. These studies and diagnostic in the sense that, they emphasis environmental causes of conflicts which are independent variables and not subordinate to what conventionally held as primary – practical, social and economic factors.

Despite certain criticism, most influential arguments in the field have concluded that, environmental stress is linked to conflict indirectly but significantly. Emperical case studies an Haity, Chiapas, Gaza, Pakistan and Bangladesh – India have concluded that, environmental scarcity of renewable resources such water, forest, fisheries and cropland give rise to a number of deleterious social efforts including economic decline, social segmentation and human migration- these social effects in interaction with other political, economic and social factors, can generate conflicts and instability. Central to this model was the notion of supply of ingenuity gap or a disparity between the solutions required to cope with environmental scarcity and, thus humen, social and institutional capital that could we mustered to provided these solutions. Thus ingenuity gap essentially undermines human adaptive responses.

Thus, the two issues emerge are,

1. Conflicts on sharing renewable resources.

2. Conflicts due to environmental degradations.

1. Interest in conflicts arising out of renewable resources sharing e.g. water resources conflict resolution has increased over last decades. (Divan 2004), and various qualitative & quantitative methods are suggested for conflict resolution. Interactive computer assisted negotiation support system (Thiessen et al), Graph model for conflict resolution (Kilgour, Hipel), shared vision modeling (Massond), Alternative dispute Resolution (Wolf), Multivariates analysis biplot (Losa) and Fuzzy cognitive maps (Giordano) presents some significant papers & case studies on prevention and resolution of conflicts over water resources.

Game theory provides a framework for studying the strategic actions of individual decision makers to develop more broadly acceptable solutions.

Author here, illustrates the utility of Game theory by discussing basic concepts of Game theory and presenting. Simple two by two water resources games.

Game theory problems are often multi criteria multi decision maker problems which are converted ultimately to single decision maker problems. In game theory, each decision maker plays a game to optimize his own objective, knowing that, other players decision affects his objective valve & vice versa .

Stable outcomes are not necessarily pareto optimal, though same theory provides more realistic simulation of stake holders‟ interest based behavior. Often this results into non co- operative behavior, even though, co-operative behaviour is more beneficial to all parties like Prisoner‟s dilemma.

Carrato et al (2005) believe that many natural resource management issues have the characteristics of Prisoner‟s Dilemma game. (231-231 K. Madeni)

Carraro et al. (2005) believe that many natural resource management issued have the characteristics of a Prisoner‟s Dilemma game: players‟ dominant strategy is not cooperative, and the resulting equilibrium is not Pareto-optimal. Similarly, most papers dealing with sharing natural resources problems have made the same assumption about the game to be the Priosoner‟s Dilemma. However, all common resource problems might not be Prsoner‟s Dilemmas (Sandler, 1992).

The conditions of a natural resources sharing problem might favor the possibility of cooperation (Taylor, 1987). Water resources games are not necessarily rival (there might be multiple users and usage by one user does not prevent simultaneous usage by other user). Thus, coordination among the parties might be beneficial to all and can create externalities. However, some water resources games can be treated as anti-coordination games in which the available resources is rival (the resource can only be consumed by one user), sharing the resource comes at a cost to users, and the resource is not excludable (it is not possible to prevent a player who does not pay for the resource from enjoying its benefits). Identifying the structure of water resource games is essential as the results can be misleading if wrong assumptions are made in conflict modeling. For instance, characteristics of an anti-coordination water game cannot be captured if the conflict is modeled as Prisoner‟s Dilemma. Bardhan (1993) believe that the literature usually jumps to the case of Prisoner‟s Dilemma in case of free-rid-ers. Sometimes, the player might not be able to reach his objective on his own. Under that condition (Stag-Hunt game) a player cooperates when the other player also cooperates and defect when the other one defects. In some common resource examples consequences of defection might be so bad that a player prefers not to defect if the other player defects (Chicken game) (Bardhan 1993). Here, two non- Prisoner‟s Dilemma water resource games, useful for understanding water conflicts, are introduced to support the fact that not all water resources games are Prisoner‟s Dilemmas.

Chicken game

In this game (Fig.4) two drivers, heading towards a narrow bridge from opposite directions, are driving toward each other. The first driver to swerve (“Chicken” out) yields the bridge to the other driver and loses. No driver entering the race wants to be the chicken, but if no driver chicken out, both drivers might suffer from the resulting crash. Being called a “chicken” is better than dying, but worse that winning, for both players do not gain anything and the fight is over protecting their prides. If their prides are more important than their lives to them, they might both die proudly ! The payoff of each player in this game can be the value of the prize at the end of the game or the utility from winning or losing the game. The higher the payoff, the more preferred is the outcome.

The Chicken game has two Nash Equilibria in which one driver loses and one driver wins, (DS, S) or (Win, Lose) and (S, DS) or (Lose, Win), which are also Pareto-optimal. The third Pareto-optimal resolution (S, S) or tie, a socially optimal resolution where the gain of each player exceeds the minimum gain in the other two possible states (3>2= minimum {2, 4}). This socially and Pareto-optimal outcome (S, S) is not a Nash equilibrium and might not occur when players make decision based on self interest.

In the Chicken game the strictly dominant strategy is to play exactly the opposite what the other player does. Similar to the Priosoner‟s Dilemma game, each player wants to get a free ride and the cooperative or agreeable mutual solution (S, S) in Chicken and (DC, DC) in Priosoner‟s Dilemma) is not stable since each player is willing to refrain from it. However, these two games differ in that if both players decide to get free ride, the resulting outcome is the worst for both players in Chicken (DS, DS) while the resulting outcome in Priosoner‟s Dilemma (C, C) is suboptimal, but not the worst for both players.

Chicken games are rare in the water resources literature as most water resources sharing problems have been treated as coordination games and modeled as Prisoner‟s Dilemma. An example of an anti-coordination water resources game is the Iran-Afghanistan Conflict on Hirmand (Helmand) River at time of the Talliban regime in Afghanistan.

The Hirmand River flows from Afghanistan to Iran and is important for agriculture in both countries as well as the survival of Hamum (Hamoun) Lake, an internationally recognized marshland in Iran‟s Sistan-va-Balouchestan province. Although there is an allocation agreement between the two countries since 1972, Iran is still struggling to receive its share from the river. The conflict between the two countries has not been resolved and the situation is sometimes exacerbated by droughts and political instability in Afghanistan. When the Taliban were in power in Afghanistan, this regime was unwilling (or could not afford) to pay the operations and maintenance (particularly sediment removal) costs for the Kajaki Reservoir in the Afghan territory. As a result Hrimad River dried up below the dam affecting agriculture and urban water supply in both sides of the border, and Hamoun‟s Lake and its ecosystem were dying. While the Afghans have responsibility to maintain the reservoir system and secure Iran‟s share of the river, since the Taliban was not doing so, the Iranians thought of fixing the system on the other side of their border. During this period, the conflict‟s structure was similar to a Chicken game (Fig.5). Both side could benefit from performing the required maintenance services. Payoffs for each country were equal to their urban, agricultural, and environmental benefits minus the maintenance cost paid. The values shown in Fig.5 are ordinal. Apparently, each side was willing to get a free ride, and spend less (minimize costs) and make more (maximize revenues). The status quo of the game, (DP, DP), in which no party would pay for the maintenance was the worst outcome, due to high urban agricultural, and environmental losses. The two equilibria of this game were (DP, P) and (P, DP) in which one party would pay the maintenance costs. The game is a Chicken game, and although being socially and Pareto-optimal, the cooperative outcome (P, P) is not a Nash Equilibrium. In this conflict, the Iranians chose to chicken out and sent teams to bring the system back to operation. Although, the final result was not ideal for the Iranians (no free ride), the cost of defection (DP) for them was so high, that they preferred not to pay (P) when they found the Afghans were willing to defect (not paying).

A good tactic in a Chicken game is to reduce one‟s options and feasible outcomes of the game by signaling intention (plans) clearly to the opponent (S) early in the game. The sent signal by a party should be strong, aggressive, and ostentatious to convince the other party that defection (DS or DP) is not the right choice. In the Chicken game one driver pretentiously can handcuff his hands behind his back before entering his car, lock his steering wheel in a straight position before the game starts, or throw the steering wheel out of the window early the game starts, or throw the other player to swerve. Other example of such behavior within a tester who has locked himself to an object, a programmed security system which explodes the property it protects if someone tries to trespass, or a nuclear doomsday device which is programmed to explode in case of invasion (“Doctor Strangelove”). In case of the Iran – Afghanistan Conflict on Hirmand, it was obvious to the Iranians that the Taliban were unwilling or unable to cooperate under any condition. The Shia Muslim Iranians had never recognized the Sunni Muslim Taliban as the legal government of Afghanistan and the two governments had no political relations. The ongoing wars among Afghan parties also made the Taliban Politically and economically unstable. The aggressive behavior of Taliban benefited the Afghans as a clear signal from the Taliban side, and the Iranians preferred to chicken out to pay for maintenance.

Unlike Priosoner‟s Dilemma in which parties lose together, the Chicken game has one winner and one loser. The Chicken game‟s structure and payoff values leave no incentive for cooperation. In water resources problems with a Chicken game structure, one might promote cooperation by increasing the penalty for defection (non-cooperation). For instance, for a conflict between two farmers over paying the maintenance costs of the pumps and irrigation channels which they both use (Fig.6a), a higher authority with a superior power (such as a farmers union or an irrigation district) can impose extra charges on farmers who do not pay maintenance costs, and so promote cooperation and prevent defection (Fig.6b). Without penalties (Fig.6a), each farmer prefers to get a free ride and not pay, leading to the system‟s demise. In that case, each farmer prefers not to defect prefers to cooperate when the other farmer defects (does not cooperate) and to defect when the other farmer does not defect. However, if defection has costs (high enough penalties) (Fig.6b), a player is not interested in getting free ride, cooperation becomes a strictly dominated strategy, and the cooperative resolution (P, P) becomes a dominant strategy (and

Stag-Hunt (assurance) game

In this game (Fig.7) two individuals who are out hunting can choose between hunting a stag together and a hare individually, without knowing the other player‟s choice. A stag has the highest payoff for both players (half of a stag‟s value goes to each hunter) but can be hunted only when both players cooperate. Instead, each player can choose to hunt a hare on his own which has a lower payoff. The Worst case for given player occurs when he chooses to hunt a stag (cooperation) and the other player chooses to hunt a hare (defection).

Table 6 reviews the characteristics of the three games introduced so far. Similar to Priosoner‟s Dilemma, Stage-Hunt is a coordination game and the two games might be confused. In both games the cooperative resolution is Pareto-optimal and the non-cooperative Pareto- inferior resolution is a Nash Equilibrium. However, unlike Priosoner‟s Dilemma, the Stag-Hunt has no strictly dominant strategy (3>2 but 1<2) and the game has one more Nash equilibrium. Unlike Chicken, in which each playe does the opposite of the other player, in the Stag-Hunt, each player‟s interest is to do exactly as the other player. Although a Stage-Hunt does not look like a dilemma, game theory finds it as a dilemma and predicts that players do not always cooperate to reach the only Pareto-optimal resolution (S, S). In practice, sometimes, player might choose not to cooperate, perhaps due to lack of trust, which results in a Pareto-inferior result (H, H) for the game. So, the game can be also called a “Trust Dilemma (Grim et al., 1999)”.

A water resources example with a Stag-structure is shown in Fig. 8. In this game two littoral countries share a lake. Each country has one river flowing into the lake. As a result of high evaporation and reductions in seasonal flows of the two rivers from upstream consumptive use, the lake is drying up, becoming salty, and its ecosystem is deteriorating. For the lake and its ecosystem to survive, both countries must increase water released to the lake by a specific amount (say 40%). Because of high evaporation, an in-crease in flow only by one country solve the problem. The payoff of each country is the environmental benefit from increasing the inflow to the lake minus the revenue lost from decreasing upstream consumptive use (even if calculation of environmental benefits in monetary values is not possible, parties still are able to rank the possible outcomes). If both countries reduce consumptive use upstream and increase release to the lake, the environmental benefits will exceed the revenue losses from reduced upstream consumption. However, if only one country increases its release to the lake, the lake‟s problem is partially solved, the environmental benefits will be minimal, and that country‟s payoff will decrease from revenue losses from decreased upstream use. The game has two equilibria, cooperative (I, I) and non-cooperative (DI, DI).

In the Trust Dilemma, if prayers trust each other, there is no risk of failed cooperation and the players will cooperate. However, in practice, non-cooperation is a risk-free strategy, leading to an out come which is not the best, but better than the worst in absence of trust to the other players. Based on this finding within Stag-Hunt structure: the rowers of a boat stop rowing or row slower to minimize their energy loss, when they suspect other rowers are not rowing effectively, although if everyone rows at the same rate, boat speed is higher; stockholders might sell their stocks individually when the company is not performing well and there is a risk of other stockholders selling their shares, although they could do better if they all keep their stocks or sell their shares together; or countries built nuclear weapons when there is a risk that other countries develop nuclear weapons, although they all agree and know that the world would be safer without nuclear weapons.

Generally, there is not tendency to free ride in a Stag-Hunt game as the payoff for non- cooperation is insensitive to what the other player does. Therefore, if one player observes signs of cooperation from the other party, he will cooperate. Similar to a Priosoner‟s Dilemma, repetition of a Stag-Hunt game can help increase trust among the parties and leading to a Pareto-optimal resolution. In the presented water conflict, however, the parties might not have a change to repeat the game many times to find if other players are trustworthy. Instead, negotiation and clear cooperative signals will be helpful in reaching the Pareto-optima resolution (I, I).

II. Second part is to illustrate, how externalities like climate change affects the security scenario, climate change reportedly (Nitin Pai, 2008) will have tremendous impact as regional security in the Indian subcontinent. Glacial melt, rising sea level and extreme weather will exacerbate online conflicts and will require India to develop military capabilities to address a range of new strategic scenarios – from supporting International co-operation, to managing a „hot peace‟, to outright military conflict.

Impact mechanisms of climate change.

According to the IPCC, the likely range of global average surface warming over the period of this century vary from 0.30C to 6.40C, depending on the model used for simulation. The corresponding average rise in seal levels ranges from 0.18m to 0.59m, excluding the impact of dynamic changes in the ice flow2 . In southern Asia, the mean annual increase in temperature by the end of the century is projected to be around 3.8oC in the Tibetan plateua and 3.30C in South Asia and 2.50C in South East Asia3 . While there is still an ongoing debate over the pace at which the temperature is expected to rise over the coming century, potential mechanisms by which the change will affect the region are clear. These are: the changes to subcontinent‟s river systems that flow from the Tibetan plateau to the Indian Ocean, and, rising sea levels and their impact on river-deltas and low-lying islands4 . In addition, a third mechanism pertinent to the this study: extreme weather – cyclones, droughts, floods etc., that do not exclusively result from global warming but are both vitiated by it and complicate our response to the disasters it causes.

Glacial recession. The glaciers on the Tibetan plateau are the source of Asia‟s biggest rivers, including the Brahmaputra, the Indus, the Sutlej and several of the northern tributaries of the Ganges that irrigate the subcontinent. Geopolitically the source of most of these rivers, except the main Ganges, lies in China. The melting of the Himalayan glaciers as a result of the rise in the earth‟s temperature will first increase the drainage through the major river systems into the ocean, followed by reduction in the their volumes once the glaciers begin to disappear. It is projected that some of the mightiest Himalayan rivers might end up as seasonal, monsoon-fed rivers like those in southern India.

Rising sea levels. The rise in global sea levels- due to the melting of polar ice caps and glaciers around the world-is expected to result in the submergence of low lying areas: including river deltas, coastlines and small islands. This places highly populated regional cities like Karachi, Dhaka, Mumbai, Kochi and Mangalore at risk. The entire country of Maldives could disappear under the Indian Ocean by the middle of the century. In addition, the coastline could advance inland across several heavily populated parts of Bangladesh, Sri Lanka, Myanmar and Pakistan (as indeed, several parts of India).

Extreme weather. In addition, climate change risks worsening the impact of natural disasters like cyclones, floods and droughts that affect the subcontinent. This could take several forms: increasing the intensity of cyclones or floods, changing the geographical area where these occur, occurring at increased frequencies and, in general, complicating adaptation efforts. For instance, a cyclone of record strength striking a river delta that is itself deluged as a result of glacial recession would transform the disaster into a much more complicated one.

Changing the dynamics of conflicts

Global climate change, by its very nature, is a trans-national phenomenon. While its impacts will not respect political frontiers, the source of climate-related problems and the hose at risk from them might well be on different sides of national boundaries. This situation is further complicated when the boundaries themselves are unclear, contested or both. As states react to climate change issues in line with their self-interests, asymmetries in risk perceptions and the existence of unresolved inter-state disputes are likely to complicate ongoing conflicts. The following table interposes the impact mechanisms of climate against the ongoing conflict dynamics in South Asia.

Impact of climate change on ongoing conflicts in the Indian subcontinent.

In light of Game theory, Global warming situation appears to have some similarities with Prisoner‟s Dilemma.

The global warming situation appears to have some similarities to the Prisoner‟s Dilemma game, as suggested by Table IV. For simplicity of argument, the options of states are narrowed figuratively to two strategies: (a) reducing CO2 emissions and (b) building sea walls. The first option is to minimize the threat by reducing the amount of global Table IV. Ranking of Outcomes of Climate Change Strategies

warming that takes place; the second is an adaptive response that seeks to limit damage caused by rising sea levels triggered by global warming.

For a typical country, the most advantageous outcome of the four would seem to occur if it invests available resources in self- defense by building sea walls, while all the other parties significantly, reduce their CO2 emissions in an effort to diminish the threat of global warming by reducing greenhouse gas emissions (Quadrant III). The amount of global warming would be somewhat lessened for that country, which in the meanwhile has enhanced its capacity to adapt to the changes that do occur. Thus, the non-cooperating country becomes a „free rider‟ that benefits from the environmental public good of less climate change that is created by other countries.

The least desirable result would occur for a country that invested heavily in cutting CO2 emissions while all others engaged in a defensive strategy of reducing their vulnerabilities (Quadrant II). The party engaging in self-restraint to the exclusion of adaptive preparations would be highly vulnerable to climate changes caused by the continuing high level of CO2 emissions of the other states. Moreover, the benefits of its sacrifice would be shared with all the other countries, including those who invested nothing in prevention.

Of the two remaining outcomes, the more preferable would seem to be for all parties to cooperate in taking decisive action to minimize global warming by reducing CO2 emissions. Cooperating to limit the amount of climate change that takes place (Quadrant I) would appear to be a less costly venture with a higher likelihood of achieving a measure of security than trying to adapt to the greater amount of climate change that would occur if most countries opt for adaptive rather than preventive strategies (Quadrant IV).

The ranking of outcomes for the paired combinations of strategies for coping with global change would thus parallel the priority of preferences for the Prisoner‟s Dilemma game. The logic of the situation would lead countries to invest what they can in sea walls or other adaptive measures leading to an outcome in which they achieve their third preference, rather than the seemingly more desirable second preference that would result if all parties agreed to reduce emissions of CO2 and other greenhouse gases, thereby minimizing the amount of global warming that takes place.

A Quadrant I outcome as opposed to a Quadrant IV outcome can be achieved through negotiations, which will succeed only if all parties believe that the others are negotiating in good faith to reach an agreement that is acceptable to all significant parties and, furthermore, that they can be counted upon to follow through on their commitments. The general absence of these conditions in the military realm has been a major obstacle to the achievement of significant arms control during the Cold War.

The Case of Ozone Depletion

The ozone depletion regime initially defined by the 1985 Vienna Convention on the Protection of the Ozone Layer and the supplemental 1987 Montreal Protocol on Substances that Deplete the Ozone Layer is a significant example of the achievement of a Quadrant I solution to an environmental security problem that was achieved through international negotiations. The Protocol, revised in 1990 and 1992, provides for progressively stricter international regulations on the production and use of ozone depleting substances, the most recent of which mandates phasing out most of these substances by 1 January 1996.” Why didn‟t the logic of the Prisoner‟s Dilemma prevail in this case?

This remarkable diplomatic accomplishment was made possible by several factors which are usually not present in the military context, two of which will be noted here. First, while human beings can take steps to reduce their exposure to an intensifying borage of ultraviolet radiation, there is no apparent defense at any cost against the much larger environmental catastrophe that is a likely consequence of a substantial thinning of the ozone layer. In contrast, armed forces and modern weapons are viewed in many countries as a viable way of enhancing security against would-be aggressor states.

Second, the consequences of a Quadrant II outcome for State A and Quadrant Ill for the other states is not as disadvantageous as in the context of a competitive arms buildup. Threats to State A‟s security may increase dramatically if it diverts substantial resources from military defense while other states are adding to their arms capabilities. Alternatively, what other countries might do to defend themselves against ultraviolet radiation poses no threat to State A regardless of whether it opts for minimizing or adapting to the problem. Furthermore, restrictions on trade in, ozone depleting substances make it unlikely that other states will derive a competitive advantage from failing to comply with the revised Montreal Protocol.

Table V. Ranking of Outcomes of Responses to the Ozone Depletion Problem

These two factors alter the priority of outcomes to the pattern suggested by Table V, in which the options have been figuratively simplified to phasing out CFCs, a preventive strategy, and using sunscreen, an adaptive one. Since there is no prospect of an effective defense against the consequences of ozone depletion, Quadrant I is the most desirable outcome for all countries because it minimizes the threat. Quadrant IV is the least desirable result in that nothing is done to limit ozone depletion. Quadrant III is less attractive to State A because of the ineffectiveness of any defensive measures it might take, and Quadrant H is less disadvantageous because there is no added threat. Thus, taking action to limit the threat results in a better outcome regardless of what the other states do. If all parties follow this logic, the outcome of their choices is Quadrant I, which appears to be what actually occurred. 12

Implications for Negotiations on Climate Change

Negotiations on a strategy for limiting the emission of CO2 began in February 1991 in the specially constituted Intergovernmental Negotiating Committee and led to adoption of the United Nations Framework Convention on Climate Change at the 1992 Earth Summit in Rio de Janeiro, where it was signed by representatives from 153 countries. The original convention was a disappointment to many for its lack of a specific timetable for reducing emissions of green- house gases. It does, however, call upon the parties to stabilize concentrations of greenhouse gases in the atmosphere at a level that would prevent „dangerous anthropogenic interference with the climate systems within a time frame sufficient to allow ecosystems to adapt naturally‟. Furthermore, it does provide for a process of continuing negotiations on additional national commitments (see Leggett & Hohnen, 1992; Parsons et al., 1992).

There are reasons for caution about the prospects for the achievement of a comprehensive international strategy on limiting global warming. First, the ozone depletion accords were reached on the presumption that substitutes for the controlled substances could be developed and produced at an affordable cost. By contrast, it is widely assumed in policy-making circles that altering energy production and use practices to the extent necessary to keep atmospheric CO2 concentrations low enough to avert a significant warming would be a very expensive undertaking that would require a massive commitment of resources, substantial economic disruptions, and significant sacrifices in life styles in the highly industrialized countries. Moreover, there is concern in poorer countries that participating in a global assault on climate change would seriously retard their economic development.

Second, whereas discovery of the Antarctic ozone hole by a British scientific team in 1985 provoked a sense of alarm, especially when it was definitively linked to human pollutants, no comparable surprise or crisis has spurred negotiations on global warming.

Moreover, there is continuing skepticism in some scientific and policy-making circles about the warnings of global warming (e.g. Singer, 1992). Moreover, while an effective defense against the consequences of ozone depletion appeared to be impossible, there is a school of thought that practical adaptive steps can be taken, at least by some societies, to cope effectively with global warming that may even be less costly and disruptive than trying to prevent or minimize climate change.

Third, while ozone depletion was generally viewed as global, peril that would have serious adverse effects for all countries, negotiations on preventing climate change have been complicated by the perception that some countries may be far more affected by climate change than others. For example, while rising seas would flood much of the agricultural land of Bangladesh and Egypt, and possibly the entire Maldive Islands, landlocked countries such as Switzerland and Austria would have nothing to fear from higher sea levels, although climate change may affect them in other ways. Moreover, some countries might envision themselves as net gainers from global warming in that the advantages of climate change, such as a longer growing season, might outweigh the costs (Glantz et al., 1990).

Finally, preventing climate change is also likely to be significantly more burdensome for some countries than others. Substantially greater sacrifices may be expected of the countries that are heavy contributors to the problem, such as those that depend heavily on fossil fuels or are economically dependent on the export of tropical hardwoods. The wealthier countries may also factor in the costs of additional economic and technical assistance that developing countries will need to minimize their contributions to global climate change.

The states most likely to adopt a defensive strategy in the pattern of the Prisoner‟s Dilemma are those that foresee fewer adverse impacts from climate changes, are confident of their capability to adapt to them, and would bear a substantial share of the cost of the global cost of preventing climate change. Conversely, states may be more inclined to support international efforts to minimize the threat of climate change if they believe they have much to lose from climate change, have little adaptive capacity, and would bear a relatively small share of the costs of preventing global warming. The United States, as well as several other highly developed countries and the transitional states of the former Soviet bloc, may fit into the former group; many of the less developed countries into the latter one, at least on the first two variables (Homer- Dixon, 1991, p. 88). Thus, the potential exists for North/South polarization on the climate change issue.

Such a polarization may not be inevitable. The costs of preventing global warming through strategies such as energy conservation may not be nearly as burdensome as is widely assumed, especially when balanced out against the costs of climate change (see Cline, 1992; Romm & Lovins, 1992/93). Moreover, governments of industrialized states may be impressed by the beneficial impacts that reduced energy production and consumption would have for lessening other environmental problems, such as acid deposition. Countries with extensive tropical forests are becoming more aware of the economic benefits of sustainable use as opposed to widespread clearing of them.

Furthermore, if evidence continues to mount on the varied secondary and tertiary impacts 0f global warming, confidence that adaptive strategies can be successful and affordable is likely to wane. Fewer states will conclude that global climate change will on balance be advantageous to them. In a highly interdependent world, even those that are less affected directly by climate change may be disadvantaged by dislocations that occur in other countries. The commitments of numerous developed countries to stabilizing and eventually reducing CO2 emissions within the decade, and the apparent receptivity of the Clinton Administration in the United States to such a goal, indicates that this type of thinking is already on the rise. Thus, there is reason for some hope, if not optimism, that the logic of the Prisoner‟s Dilemma will not prevail and that a global accord can be reached that will substantially limit climate change.