The Environmentalist 18, 77±86 (1998)

Environmental issue area and game theory HEIKKI O. KUISMIN* Kotikonnuntie 3 E 67, 00940 Helsinki, Finland

Summary The impact of international cultural, political, social, military, scienti®c, environmental and economic events has never been more vividly felt than today and the `challenge of global change' has been created. Today there are many pressing regional and global problems. The planet is under stress and so are various scienti®c paradigms. To solve the new problems, international multilateral action is needed and increasingly sought. Mutual and not sporadic cooperation is the most prudent policy, the most effective means to attain set environmental goals such as clean air, clean water and clean soil. As to cooperation, national actions taken by state authorities cannot abolish any of these problems, but regional or preferably global, international multilateral action must be taken to meet the new demands. Responsibility for future generations is the key. Which is preferred: growth at any cost with widespread ecological destruction or sustainable development with cleaner water, air and environment? What is prudent policy? Cooperation or competition? Do we need long-term planning instead of short-term planning and gains? These are the key questions which this paper seeks to answer. Introduction All persons are enveloped in the net of world politics. The impact of international cultural, political, social, military, scienti®c, environmental and economic events has never been more vividly felt than today; the `challenge of global change' has been created. A rede®nition of what constitutes national security is, therefore, in order. The traditional agenda has included highly political, classical issues such as nuclear proliferation, the balance of power, the arms race, high-intensity and low-intensity con¯icts and subversive national liberation movements, all of which have receded into a less dominant position on the security agenda, although intervention has not disappeared from the international scene nor will it in the future. The information war ± a nascent, perhaps embryonic, art ± is the latest form of this ancient art. It has become evident with the end of the Cold War that the security agenda in the 1990s and beyond will be signi®cantly di€erent from the one to which humans have been accustomed. Previ* Mr Heikki Olavi Kuismin is a licenciate in international politics. At present he is in the process of completing a PhD thesis `Changing Security Agenda, States, People and the New External Threats' at the University of Helsinki, Finland. He has published many articles on the environment and world politics in general.

0251±1088 Ó 1998 Chapman & Hall

ously neglected and ignored questions and issues have arrived on the agenda. Now, at the dawn of the twenty-®rst century, there are many pressing regional and global problems, such as population explosion, world hunger and refugee problems, issues of racial discrimination, human rights, international economics, religious and ethnic groups, depletion of natural resources, international crime, AIDS and diculties in the ways of promoting and de®ning the rights of women. The planet is under stress and so are various scienti®c paradigms. There are truly global environmental problems or threats ± both external and internal ± such as pollution, disposal of nuclear and other waste, radioactive fall-out and such accidents, which can a€ect the lives of millions of people. For example, Greenpeace (1996) stated that some 9 million people su€ered from the Chernobyl incident of 1986 and that approximately 375 000 people were evacuated from the polluted area, which covered 100 000±160 000 km2. It has also been assessed that the cleaning of the contaminated land will last till 2015 and will cost approximately US$300 million. In addition the greenhouse e€ect, acid rain, ozone layer damage, deserti®cation, deforestation in general and the mass destruction of the rain forests in particular are new important external threats with which states must cope. A

77

Kuismin

plethora of books dealing with the new threats have been published in recent years (e.g. Tomasek, 1987; Mathews, 1989; Luard, 1990; Martin and Romano, 1992; Kennedy, 1993). Much damage has already impacted on Earth's DNA bank, that is, its fauna and ¯ora. Many species have become extinct and more will rapidly follow without positive intervention. These issues and many more have come within the purview of the political scientist in recent years. These many and varied new internal/societal (emanating from within society) and external/ systemic (emanating from without) negative factors and forces, along with the more traditional threats, have impaired states' performance. It is increasingly dicult for nations to perform their basic functions e€ectively. Today, people are living in a complex, interdependent and interactive world in which a harmful situation abroad can a€ect the welfare, security and behaviour of numerous macro and micro actors, indeed, perhaps all of them (e.g. ozone depletion; global warming may even bene®t some actors like the rich societies in the north). The expansion of knowledge and information, although still imperfect, makes many events seem less random and more speci®cally attributable. The spewers of acids, greenhouse gases and CFCs are identi®able actors. The steady uncovering of the possible cause±e€ect relations puts ecological issues more and more into the human and, therefore, the political arena (Buzan, 1991). Many environmental factors and forces can and already do endanger the lives and health of people. Today, these factors are important in setting political priorities. International cooperation International multilateral action (i.e. actions taken by several state and non-state actors; there are good reasons for not using the term multinational, since rarely do nations act, nor is this term sucient to describe the complex world and the myriad of transactions now taking place across state borders) is needed and increasingly sought. Mutual and not sporadic cooperation is the most prudent policy, the most e€ective means to attain set environmental goals such as clean air, clean water and clean soil. As to cooperation, national actions taken by state authorities cannot abolish any of these problems, but regional or preferably global international multilateral action must be taken to meet the new demands. Political action has traditionally been undertaken within national political systems. Individuals and groups have competed for power within the state because it has been state authorities that have taken the actions ± economic, political, social and environmental ± that have determined citizens' welfare. Old-style political activity of that sort is no longer relevant to the needs of modern citizens. The decisions that matter most today, 78

that determine security and economic welfare, promote social justice, protect human rights and safeguard the environment, are reached not by national but by international authorities. E€ective political action today, therefore, must be international or multilateral action, directed at such authorities (Luard, 1990). The development of environmental problems across state borders ± or state frontiers (on the distinction, see Glassner, (1996) ± is the most important international externality that engenders signi®cant multilateral cooperation. Global power struggles and irresponsible usage of the environment have become ever-more negative sum games, from which all or most of the national and local governments began to realize from the mid-1980s that they and their societies had nothing to gain and everything to lose. The number of ecotreaties has rocketed since the 1980s and numerous intergovernmental organizations have been established to deal with environmental matters (Caldwell, 1990). The birth of an international environmental regime is being witnessed. The international conduct of states is restrained by moral, institutional and legal bonds as far as environmental issues are concerned. An international regime refers to the sets of governing arrangements a€ecting relationships of interdependence (Keohane and Nye, 1989). E€ective administration and decision making is dependent upon international regimes, sets of implicit or explicit principles and networks of formal or de facto norms, rules and procedures, frequently unstated, but which regularize the behaviour of major actors. They control its e€ects and guide and facilitate international decision making and multilateral action. Principles are beliefs or facts, causation and rectitude, while rules are speci®c prescriptions for action. Niebuhr (1948, quoted in Dougherty, and Pfaltzgra€, 1981) wrote that `[O]ur problem is that technics have established a rudimentary world community but have not integrated it organically, morally or politically. They have created a community of mutual dependence, but not one of mutual trust and respect'. In the international cooperative interaction game the pay-o€s should bene®t the players, i.e. the actors can gain more as a unit. The clashes between short-range and long-term interests, costs and gains are the principal issues. Cooperation, de®ned in terms of conscious policy coordination, is necessary for the realization of mutual bene®ts: the players must prefer mutual cooperation to mutual defection. Can the other actors be trusted? What are the long-term intentions of the other players? For how long will they cooperate? What are they really up to? How credible are their threats and promises? These are some of the relevant and dicult questions decision makers must ponder. Psychologists have separated temptation, the desire to obtain the largest pay-o€ by being the only defector, from The Environmentalist

Environmental issue area and game theory

a two-player game defection is often more easily detected than in an n-person game, e.g. with 150 players acceded to an environmental convention. Cooperative games are games with enforceable commitments and with free communication between the actors. The latter condition is often met, but the presence or absence of free communication is only of secondary importance: what counts is the possibility or impossibility of enforceable and binding agreements. The agreed pattern of adjustment relies on voluntary patterns of compliance, somewhat reinforced by public opinion, rational self-interest (true, enlightened or healthy (Gabriel, 1994) selfinterest) in the form of cleaner air, water and soil within state borders and pressure from some international non-governmental organizations (INGOs), in particular Greenpeace. It remains to be seen whether, for example, some of the larger countries in the `south' (i.e. middle powers such as India, China, Indonesia, Brazil and Mexico) will in the end reject cheaper CFCs and impose more strict national laws and replacement technologies in their rush to introduce refrigeration and aerosol spray devices. The experience with ozone depletion will be an important test of whether or not sovereign nations have the capacity to deal successfully with aspects of the environmental challenge. In world politics actors have often defected from cooperation, violating treaty provisions. In world politics monitoring treaty compliance and veri®cation is often dicult. If the game is played, therefore, as a non-cooperative one, the outcome will be (in theory, not necessarily in real life) the equilibrium point Ei,j, i.e. mutual non-cooperation and is often called the non-cooperative solution. If the game is played as a cooperative game, then the outcome will be the non-equilibrium point strate-

mistrust, that is the fear that the other actor would like to be the lone defector. They found that temptation is a more likely source of non-cooperative behaviour than is mistrust. Often, actors can gain through functional cooperation. In general, unilateral defection has been preferred to unilateral cooperation, but the actor choosing a unilateral cooperative strategy gains more in the long run than the actor defecting from cooperation. It does not really matter whether other actors can be trusted or not, because unilateral cooperation is greater than unilateral non-cooperation which is greater than mutual non-cooperation. In all circumstances, actors should choose cooperation instead of non-cooperation. In game theory in general, there is an incentive to defect, although in reality mutual cooperation would yield a higher pay-o€ (e.g. cleaner air, soil and water) than mutual non-cooperation resulting in a contaminated environment in both or all countries. One precondition for cooperation is that the actors are more or less convinced that mutual, often rather vague long-term gain is possible in the non-zero (variable) sum environment. Cooperative game means that the players can make fully binding and enforceable commitments and fully binding agreements, threats and promises which must be implemented if stipulated conditions arise. In non-cooperative games this is not the case. In real life, what makes commitments (almost) fully binding is a law-enforcing authority. There is no such authority, an international Leviathan, in the contemporary world system. However, in many cases, prestige considerations (fear of losing face) may instead operate. In a cooperative game the players can agree on any possible combination of strategies since they can be certain that any such agreement would be kept. This makes cooperative behaviour more likely. In Actor Y

Actor X

Actor X

Actor Y

Actor Y

Actor Y

A

E

A

E

A

E

A

E

A

3,3

2,4

3,3

1,4

2,2

1,4

1,1

2,4

E

4,2

1,1

4,1

2,2

4,1

3,3

4,2

3,3

Chicken

Prisoner's dilemma

Deadlock

Deadlock type

Actor Y

Actor Y

Actor Y

Actor Y

A

E

A

E

A

E

A

E

A

2,2

3,4

4,4

1,3

4,4

2,3

1,1

3,4

E

4,3

1,1

3,1

2,2

3,2

1,1

4,3

2,2

Leader

Stag Hunt

Harmony

Hero

Key: (x, y) = rank of row (X), rank of column (Y). 4 = best, 3 = next best, 2 = next to worst and 1 = worst; A = cooperation, E = non-cooperation (Nash equilibria in bold type). Fig. 1. The principles of game theory.

79

Kuismin

gy pair Ai,j, i.e. both actors cooperate. This is called the cooperative solution. More generally, the solution of a non-cooperative EE game must always be a Nash equilibrium point where each actor's solution strategy must be a best reply to the other actor's solution strategy, (Fig. 1). As to rational choice, prisoner's dilemma and `free rider' problems in environmental international cooperation suggest that the rich Third Wave countries in the north (as a `bloc', actor X) should use strategy Ai even if they knew that less-developed countries (LDCs) would defect from using cooperation strategy, Ej. This is because 70% of all emissions of greenhouse gases accelerating global climatic changes come from Third Wave (To‚er and To‚er, 1994) and Second Wave industrial societies, from the countries in the rich north. These societies should therefore assume the main burden of reducing the global level of emissions, not the countries in the poor south. Rich countries can a€ord for them to be free riders because only 30% of all emissions come from those countries. In short, because of the 70:30 ratio, the possible freerider problem should not be allowed to prevent large group activity, concord and conventions from taking place. It is assumed that the actor has complete and full information about all the parameters de®ning the game, and about all the variables determined before the beginning of the game (e.g. actors have sent reasonably clear signals of their future behaviour). These variables include the actors' payo€ functions, the strategical possibilities for each actor and the amount of information each actor has about all of these and possibly other variables. Game theorists speak of a game with perfect information if the actors have complete information about the moves already made in the game, including the personal moves made by the individual actors and the other moves decided by chance. Thus, perfect information means full information, requiring e€ective monitoring and veri®cation systems when free riding becomes dicult, about all the game events which take place after the beginning of the game. In a non-cooperative game only self-enforcing agreements are worth making since only they have any real chance of implementation. The self-enforcing agreement is called an equilibrium point. A given strategy of a certain actor is called a best reply to the other actor's strategies if it maximizes this actor's pay-o€ so long as the other actor's strategies are kept constant, which they rarely are. A given combination of strategies is called an equilibrium point, if every actor's strategy is a best reply to all other actor's strategies. The prisoner's dilemma can be a two-person non-cooperative game and it is possible to assume (see Fig. 1) that no enforceable agreements can be made. However, as to the dyadic relationships and interactions in general, whatever may be the attractions of simplicity and elegance in theory building, there is 80

always the counter-attraction of a model retaining formality even while incorporating further complexities. In world politics, the appropriate formal image is a number of kinds of actor simultaneously engaging in a series of multiplayer reiterations and not single-shot, non-variable sum games (Buzan et al., 1993). Some scholars have used the game of prisoner's dilemma to explain why in many countries higher chimneys have been built in recent years. It is assumed that actor X does not (albeit does in reality) pollute its own environment and air, choosing non-cooperation; this EA strategy yields actor X its best outcome x4, while the countries at the receiving end (as a `bloc', actor Y) obtain their worst outcome y1 since actor X pollutes their environment. Actor X hopes that actor Y will not follow suit; if it did, actor X's outcome would then be the next to worst ± x2. Cooperation would yield x3, the next best outcome with certainty. There are two conditions to be met in prisoner's dilemma namely U …y4 † ‡ U …y1 † < 2U …y3 † and U …x4 †‡ U …x1 † < 2U …x3 †; U refers to utility. The linear payo€ for actor X is Ux …Ax ; Ay † ˆ Ax ÿ 2Ay . Players have a temptation to defect from cooperation (EA, AE) to obtain their best result (i, j4) at the expense of the other player, when j, i1. Let Ax present actor X's level of cooperation measured on a scale from 0 to 1. Here, 0 represents full cooperation while 1 represents full defection. Similarly, Ax 2 ‰0; 1Š represents actor Y's level of cooperation. Temptation is the rate of change of actor X's pay-o€ with an increase in its defection level, while fear is the absolute value of the change in pay-o€ su€ered by actor X when actor Y increases its level of defection. So, temptation ˆ aUx =aAx ˆ 1 and fear ˆ aUx =aAy ˆ 2. This game is prisoner's dilemma de®ned on the unit square [0, 1] ´ [0, 1]. In the repeated game the two actors simultaneously choose their level of cooperation/defection (A, E) at each point in time t, thus determining an outcome At ˆ …At ; Aty † in the square. If the actors are concerned with the future consequences of their decisions and actions, a natural representation of how the two actors value their relationships is the discounted pay-o€ stream Vx (for actor X, with discount factor xx , at time t): Vx …t†

00 X sˆ0

t‡s xsx  Ux …At‡s x ; Ay †:

It is possible to arrive at game theoretical ratios intuitively and it is not always easy to assess how one outcome can be ®ve times as good as some other outcome. Certain outcomes can be represented in negative terms. The pay-o€ values are often assigned in a purely arbitrary manner, merely to facilitate the illustration of a point. There is no need for an absolute value to be assigned for each outcome, but merely that an ordinal ranking …x4 > x3 > x2 > x1 † of utilities is established. Moreover, the strategies may consist The Environmentalist

Environmental issue area and game theory

of fairly complex plans and yet be designated simply (in doing so, it is possible to understand complex real life situations better by simplifying them and capturing their essential characteristics) strategy A, strategy E or strategy N for each player. Thus, in mathematical theory, both strategies and pay-o€s are treated abstractly. Each matrix contains the pay-o€ which each actor receives when they choose one of the two (E or A) strategies that converge at that point. Actors are assumed to be rational, striving for the attainment of their preferred outcome. As far as the pay-o€s of both chicken and prisoner's dilemma are concerned, mutual cooperation (AA, AA) yields only i, j3, the next best outcome, whereas defecting from cooperation (EA, AE) would give in both games i, j4, respectively, i.e. the best outcome. With these remarks as a prelude, the present paper turns to environmental problems, the new important internal and external threats that a€ect citizens' welfare everywhere. As far as environmental issues are concerned game theory has proved a bad ®t because the preference orderings of the existing games have been irrational as to real life situations. Game theory has been widely employed in economics but, as for the applicability of games to world politics, the `political' is not perfectly interchangeable with the `economic' or the `psychological'. There are important di€erences between political decisions and those made by private companies and private citizens. Before turning to the empirical problems, the paper ®rst makes an ontological point. International actors are not free agents who can do whatever they please. Other actors, both domestic and foreign, with their possible negative responses constitute an important countervailing power, impairing actors' goal attainment. That is why there is, as a consequence, extreme scepticism over attempts to develop a theory of action expressed only in terms of intentions, beliefs, goals and desires of individual actors. The outcome of many actions taken by one actor tends to be contingent upon actions taken by other actors. Such situations are characterized in terms of strategic interaction. One of the reasons why game theory has become so popular as a methodology in social sciences (at least in the USA) is because it reveals in straightforward and clear terms how the choice by one actor a€ects the choices of other actors. A game theoretic (for instance, the very popular 2 ´ 2) matrix describes how the choices of two actors or groups of them intersect and illuminates how the structure of the situation they ®nd themselves in constrains their choice of action. The matrix also reveals that no rational actor, operating in a structure where two decisions arrived at independently intersect, can fail to take the possible actions and negative responses of the other actors into account. However, some actors

(the `losers') may fail to take into account other actors' choices, perhaps unintentionally (there might be a `Sproutian' gap between the psychological and operational environments, often with disastrous consequences) or intentionally, when the actor is either powerful enough to engage in such an attempt or taking risks (showing often greater daring than judgement; however, from time to time these kinds of attempts have paid o€). Environmental issues and game theory Not all games and their preference orderings and the rationality associated with the choices are congruent with reality and not all games can be applied automatically per se to all branches of social science. This `bad ®t' would most likely yield poor results. In politics, actors exchange verbal and non-verbal negative (designated E, which refers to `enmity') and positive (A, referring to `amity') signals. In game theory, actor X can respond to the negative or positive signals sent by actor Y by sending negative signals, xEy or by sending positive signals, xAy. Actors interact and transact with one another, but they are not simple tennis matches, action-reaction patterns, nor can complex processes be cast in binary, `either±or' terms. All processes end in due course. The causes underlying their termination or stand-still are many and complex. In reality, there is a third alternative, inaction, when xEy/xAy ˆ 0. After inaction, the process may end, as it can after xEy ! yAx, when actor Y responds by sending positive signals to actor X's enmity signals. In order to make predictions about the likely outcomes of a speci®c game, one can use the standard Nash solution to ®nd equilibria. This solution requires the players to evaluate each situation in which they may ®nd themselves with respect to what they may achieve on their own by unilaterally changing their strategies to ameliorate their situation (Aggarwal and Allan, 1992). It is possible to arrive at the game theoretical ratios intuitively. It is not easy to assess (if nonmeasurable, intangible things are involved) how one outcome can be ®ve times as good as some other outcome; this is not an issue here. Note also that outcomes may be represented as negative terms, but not in this text. The pay-o€ values are often assigned in a purely arbitrary manner, merely to facilitate the illustration of a point, e.g. i1, j1 (this paper designates actor X as i and actor Y as j), resulting from mutual non-cooperation in the game of chicken. To give but one example, nuclear war would be the worst outcome in reality, as is assumed in chicken, too. Hence, there is no need to assign to the negative, non-cooperative EE (xEy $ yEx)type of interaction, a pay-o€ value of 1 000 000 or any other value. No absolute value need be assigned for each outcome but merely an ordinal ranking of utilities has to be established. 81

Kuismin

Moreover, the strategies may consist of fairly complex plans and yet be designated simply as strategy A (cooperation) or strategy E (non-cooperation). Each matrix contains the pay-o€ which each actor receives when they choose one of the two (E ˆ non-cooperation and A ˆ cooperation) strategies that converge at that point. As usually happens in game theory, the actors engaged in the game are assumed to be rational, striving for the attainment of their preferred outcome. A visual examination of the matrices (Fig. 1) suggests that the games of leader, hero, prisoner's dilemma, chicken, deadlock and deadlock type are not apt for an evaluation of rational environmental policies. For example, deadlock type is like deadlock where both actors' dominant strategy leads to an outcome of no consensus: the `best' result for one actor is the `worst' for the other. The game of leader bears some similarities to chicken, in which the likely outcome is asymmetric as one actor prevails over the other. The game of leader has often been applied to model situations of coordination, whereas chicken is used to represent cases where one actor imposes its will on the other through reckless and irrational behaviour (e.g. deterrence is based on irrational future behaviour). In the latter case, the asymmetry between actors is much greater and the relative gain greater for the victorious actor in chicken than in leader. As to environmental issues, the games recounted above have preference orderings which do not ®t in with the author's theory of prudent environmental policies. As to the pay-o€s in chicken and prisoner's dilemma cooperation (AA; xAy $ yAx) would yield only i, j3, the next best outcome, whereas non-cooperation (EA $ AE; xEy ! yAx) results in actor Y being the `sucker' and vice versa (xAy ! yEx), actor X being the sucker. In all games one of the actors engaged in an AE game is a sucker, losing and getting a worse outcome. Thus, because AA [(xAy $ yAx) ˆ i, j3] < EA/ AE [(xEy ! yAx/xAy ! yEx) ˆ i, j4] for one of the actors, these games with their `rational' moves (c.f. preference ordering) are incongruent with reality, particularly as far as environmental issue area is concerned. Game theorists distinguish the outcome of a game (win i, j4, lose i, j2 (in chicken) or i, j1 (in prisoner's dilemma) or draw i, j3 from the pay-o€, the value attached to the outcome by an actor (the relationship between pay-o€ and motivation is critically important but it is dicult to establish). The outcome of EA (i defecting from cooperation) in the short-run might be i4, j2 (or i4, j1 in prisoner's dilemma), but not in the longrun; the reverse is the case with acceding to international environmental treaties. Yet it is dif®cult to conceive how the lone defector actor X or i could gain by polluting its national environment, except that its domestic private companies would gain alleged economic competitiveness. 82

A word on rationality In game theory, one question arises: `What is rational behaviour?' It is not always easy to say what is rational or irrational in certain conditions, nor is it easy to tell who is rational and who is not. Rational choices often depend on the context, time and place plus the issue (and stakes) in question. In addition, the preference orderings of the actors a€ect their choices; a certain move may seem absolutely irrational to outsiders, but be utterly rational to the player. Calculations of unacceptable damage made in di€erent countries yield di€erent outcomes since people have widely differing beliefs, personal histories, cultures, value systems and conceptions of national security and international objectives. As can be seen, rational is one of the many value-laden terms used in the study of world politics. To put it succinctly, `rationality' is very much a context-bound concept (Elster, 1986; Nicholson, 1992) as is the term `terrorist', which is also a value-laden term. Many extremists (amongst the `greens') see big companies and their leaders as `ecoterrorists', whereas the leaders of the companies conceive of these extremists, some of the members of Greenpeace for instance, as ecoterrorists. The new ecological threats tend to upset the two Hs (called `national security' and `welfare of the citizens' homeostatic variables, i.e. the principal ends or functions states have) to national security, like economic and military ones. They can damage the physical base of the state, perhaps even to a sucient extent to threaten its very idea and institutions. Traditionally ecological threats have been seen as random, something `given' or inevitable, part of the natural conditions of life and, therefore, more a matter of fate than an issue for the national security agenda. Many acts of God such as hurricanes, plagues, ¯oods, night frosts, earthquakes and droughts might in¯ict even war-scale damage on a state and can lead to mass immigration and dislocation of people exacerbating a refugee problem. These have been seen as part of the struggle of humans against gods and nature, whereas national security issues arose much more from the struggle of humans with each other. For most of history, the environment has been a relatively constant background factor rather than an issue in its own right, through there have been many environmental theories of world politics. However, with the dramatic increases in the diversity, scale and pace of human activity in the nineteenth and twentieth centuries and with the rapid expansion of knowledge about the planetary ecosystem the environment is no longer a background factor. Core corporations have consistently argued that in a highly competitive world environment greater regulation threatens higher productivity and business con®dence. As a consequence, decision makers, fearful of politicizing their ®nancial The Environmentalist

Environmental issue area and game theory

and economic decisions, have refrained from extending the scope and ecacy of state intervention, thereby limiting, although indirectly, the reach and e€ectiveness of international organizations (Camilleri and Falk, 1992). Many more or less unsuccessful United Nations conferences are also indicative of this trend. Interests and values do clash, seeming incompatible in many countries, irrespective of their level of development. The principal question is what is more important: clean air, water and soil or the alleged competitive advantage of the nation (or rather of its big companies)? Besides, there are several more negative factors a€ecting competitive advantage much more than green technology and restrictions. New technology would be only a slight burden compared to factors such as high wages, long distances and high national currency. The World Commission on Environment and Development has de®ned sustainable development as a political process of change. Responsibility for the future generations is the key here. Which is preferred: growth at any cost with widespread ecological destruction or sustainable development with cleaner water, air and environment? Environmental policies and priorities will legitimately and inevitably vary from country to country and from one community to another. This re¯ects the immense variation existing in the physical world, in the nature and scope of environmental challenges and in the di€erent perspectives on environmental problems that come from di€erent levels of economic development. There is certainly a good deal of room for positive reinforcement between di€erent goals and priorities, between the maximum preservation of the natural environment and the pursuit of continuously high levels of economic development and between rapid economic growth and the protection of traditional cultures or improvements in equity and social justice (Hurrel, 1995). During the Reagan and Bush administrations a political battleground for competing priorities was staked out in the USA, where several grandiose plans for big technology and science were proposed while simultaneously reducing or withholding funds for many aspects of environmental protection. In 1990, President Bush said that the USA was going to plant the US ¯ag on Mars by the year 2019 (the programme has now been cancelled, because it would cost well over $100 billion), whereas Reagan espoused the strategic defence initiative (SDI), commonly known as `Star Wars'. During their tenures, the USA failed to sign any important environmental agreements. During the Clinton tenure the situation has not changed, although the US Vice-President Al Gore is a keen advocate of prudent environmental policies. However, one man cannot win the battle. US conservatives deem economic growth more important than the environment. The Republican Party, along with many national governments,

resorts to sovereignty discourse and justi®es its politics as designed to restore national economic competitiveness and national sovereignty, which is one variable underlying US behaviour in various international environmental conferences, at which US participants have not gone along with the strict environmental recommendations. The situation in the USA is not any better in the mid1990s: according to the US NGO League of Conservation Voters the year 1995 was, as far as environmental issues were concerned, the worst in 25 years, for example no new signi®cant environmental laws were adopted in 1995±1996. Porter (1990) of the Harvard Business School has established himself as the world's leading authority on competitive advantage. He has explored ± at a time when economic performance rather than military might will be the index of national strength ± what makes a nation's ®rms and industries competitive in global markets and propels a whole nation's economy to advance. In his analysis of ten leading nations, namely the USA, the UK, Sweden, Denmark, Japan, Italy, West Germany, Korea, Singapore and Switzerland, Porter (1990) studied the patterns of (mainly Second Wave at the time) industry success as well as the company strategies and national policies that achieved it. These Organization for Economic Co-operation and Development (OECD) members include the three leading industrial powers as well as other nations intentionally varied in size, government policy towards industry, social philosophy and geography. What can be inferred from his seminal work is that strict international environmental treaties and domestic laws have but little impact on a nation's overall economic competitiveness ± there is a host of other more important negative factors than restrictive environmental laws a€ecting a nation's competitive advantage. Green technology, ecotreaties and restrictions are but `drops in the ocean'. For instance, new technology would be only a slight burden compared to factors such as high wages, long distances and high national currency. It can be readily argued that preventing pollution is a prerequisite to the satisfaction of basic human needs, most notably citizen's health and general welfare. Who wants to live at a dumping ground, drinking contaminated even poisonous water or breathe unclean air that makes one cough blood and get sick? Nobody does. Since private citizens, the former victims of world politics, have more say in domestic and world politics today, it is possible to take as their point of departure the assumption that the environment is an important factor in setting political priorities. In addition, the initial assumption can be made that actor Y prefers sustainable development and a clean environment to the policies of economic growth at any cost after performing a long-run costs-togains ratio analysis. However, many economic factors and short-term planning seem to impair 83

Kuismin

rational long-term calculation, in particular in the rich north. It seems clear that the former goal can be denoted as G4; it is regarded a fundamental part of the two principal goals, namely G1 and G2. The economic goal, then, recedes into a less dominant position, G5 in the preference ordering. In many countries judgement in this area is often in¯uenced more by scienti®c analysis than political preference. Thus, a player choosing AE after `AA and AE' or yAx $ xAy ! xEy is not a sucker at all (unlike in all the other games referred to above, even in Keohane's harmony and Rousseau's stag hunt), but a `winner', the most rational player. It is almost a truism that world politics cannot be de®ned in zero-sum terms, in other words that any gain for one party must be made at the expense of another or others. As to the environmental issue area, by contrast, a cooperative situation (AA or EA/AE) is one in which the behaviour of one actor confers bene®ts upon another. Fortuitously if that is unintentional, the preference ordering recounted above can be called `paradox'. To clarify some of the things discussed above, the author represents the `paradox' in a simple matrix. In doing so, the complex real life situations can be better understood by simplifying them and capturing their essential characteristics. Policy decisions by state and non-state actors in interdependent situations can best be examined with the help of game theoretical models, but, albeit these models are very useful in analysing strategic interactions, they have one signi®cant limitation: game theory does not advise which speci®c game structure to use in a given empirical instance. In other words, a theory is required of what kind of game to use as a model for di€erent cases. The analyst's theoretical approach tells them which particular game structure to use in a given situation, starting from the key situational determinants of decision makers. In paradox this is not actually a `game' in the classical game theoretic sense, since there are no tensions; the author employs a game theoretic matrix just to illuminate succinctly and clearly how a sucker can actually win) two actors or groups can rank the four outcomes from best to worst, which is illustrated in Fig. 2. The actors in each game are assumed to be able to choose between the strategies of cooperation (A) and noncooperation (E). The choices of strategies by each player lead to four possible outcomes, ranked by the actors from best (4) to worst (1) (Brams, 1985). The ®rst number in the ordered (e.g. x4±x1) pair that de®nes each outcome is assumed to be the ranking of the row player (actor X ˆ i), and the second number the ranking of the column player (actor Y or j). Unlike the other games dealt with above, the starting point here is the strategic pair EiEj, mutual non-cooperation as is the case in reality, too, since all states pollute their own and many of them 84

also other actors' environments. To give two examples; in prisoner's dilemma the prisoners would be in jail in the EiEj situation and dead in chicken; they could not improve their welfare after EE, but actors engaged in paradox can and should (game theory is normative in character and some policy prescriptions can be o€ered) look for improvements. The preference ordering in paradox is based upon the conception of a clear long-term calculation of interest and the avoidance of shortterm and self-defeating policies of intransigent `no change', even though these may suit the interests of those temporarily occupying decision-making roles. Actors must begin to view their mutual problem as a positive-sum game in which actors can gain in the long-run from collaborative environmental problem solving. Resolution will come about because of (long-term) self-interest, based on the realization that the costs of continuing the pursuit of the current policies with twisted (or unreal) self-interest outweigh the bene®ts gained by ®nding a solution. While such a set of policy prescriptions may appear to be over-optimistic and even naõÈ ve, the practical and intellectual barrenness of conventional politicoeconomic approaches is such that any alternative needs to be taken seriously as a guide to action for success or even survival of the mankind. People should be more concerned. From the matrix below can be inferred one selfevident policy prescription: cooperation should be the option in all circumstances. Any actor choosing cooperation will come out on top. This optimism requires the trust and information of other actors' true (enlightened) motives and intentions (c.f. the principle of bona®de in domestic law or pacta sunt servanda in international law). To put it succinctly, actors must trust one another and rely on their rationality and good intentions in the face of common and grave threats. In order to make predictions about the likely outcomes, this paper uses the standard Nash solution to ®nd equilibri-

Fig. 2. The `Paradox' matrix. Key: (x, y) ˆ rank of row (X), rank of column (Y). 4 ˆ best, 3 ˆ next best, 2 ˆ next to worst and 1 ˆ worst; A ˆ cooperation, E ˆ non-cooperation (Nash equilibria in bold type).

The Environmentalist

Environmental issue area and game theory

um. A set of strategies constitutes a Nash equilibrium if no actor's welfare can be improved by a unilateral change of strategy by that actor (e.g. from AA, xAy $ yAx, to AE, xAy ! yEx or vice versa); the concept of equilibrium is used here in this sense and is not intended to mean balance or stability. In paradox, in a one-shot game, the Nash equilibrium strategy is for both actors to cooperate, that is AA or xAy $ yAx, a positive `tit for tat' (Bueno de Mesquita and Lalman, 1992). However, most likely, as has very often been the case in world politics, breaches of trust would occur, since many actors would adopt opportunistic and myopic carpe diem policies, irrespective of its irrationality, e.g. actor X regards economic growth more important than abiding to the international environmental agreements. As for actor X's ®rst move, negative non-verbal stimulus yAx ! xEy (c.f. the concept of an asymmetric game), the outcome of the lone defector X would be i2 instead of i3, the next best outcome (should actor Y make the ®rst negative move) or i4, the best outcome after continuing a cooperative AA game (i.e. xAy $ yAx). It is not, then, rational to defect from cooperation, but it is rational to defect from non-cooperation (i, j1 ! i, j3, i.e. the actors' pay-o€ would improve from worst to the next best pay-o€), as simple inspection of the matrix reveals. Should both actors play EE (xEy $ yEx), then EA would mean improvement in actor Y's welfare, from j1 (polluting the state's territory) to j3 (not polluting the state's own territory). This, of course, should be a great incentive to continuing AA strategy, to the constant exchange of positive verbal signals, such as speeches given at the UN Environmental Conferences and non-verbal signals (actions), xAy $ yAx. The game of hero (in which Saddam Hussein was engaged in 1990±1991; in addition international terrorists, including `eco ones') along with deadlock type, is unrealistic in an environmental sense since cooperation in these two games gives the worst (i, j1) outcome. Stag hunt seems better (mutual cooperation yielding the best outcome, namely i, j4 for both actors), while harmony ®ts best, since in it the loser, that is the non-polluter taking care of its environment, obtains j2, while in stag hunt the outcome is only j1 for actor Y after an EA-type of interaction. Here actor X moves ®rst (xEy ! yAx). In reality, assuming that ecologically sound decisions are preferred to unsound ones, (which are often decisions which yield economic pro®ts) the arguments for rational preference ordering must be accepted. The non-polluter should obtain j4 or at least j3 for its ecologically sound move after an EA interaction (again, actor X moves ®rst: xEy ! yAx). Thus, with a stag hunt ordering actor Y bene®ts less by its ecologically sound policies, i.e. not polluting its own territory (i3, j1 after EA interaction), than is the case in harmony (i3, j2).

As for rational ecological decision making, harmony, with its preference ordering, seems to be the game that is more congruent with reality than the other games. The problem is, however, that even in harmony actor Y gains more by polluting its own territory (which the actor does irrespective of `high chimneys') than not doing so (j3 and j2, respectively). This paper has provided a new matrix with rational preference ordering to overcome the de®ciencies in the games recounted above. Most of the ideas presented here have been arrived at deductively and independently through the application of simple logic and common sense, though there is ample empirical evidence to support the conclusions. To give one example, a person's own territory is usually the ®rst to su€er and much more severely than that of the neighbours, therefore AE interaction (xEy ! yAx) should yield i2, j3 to be congruent with real life situations; here the EE-type of interaction (xEy $ yEx) gives the worst outcome i, j1 for both actors. Thus, the preference ordering of moves, as far as actor Y is concerned (actor Y moving ®rst), is AA (that is, yAx $ xAy) > AE (yAx ! xEy) > EA (yEx ! xAy) > EE (yEx $ xEy). As can be inferred from the foregoing, actor Y is not the sucker but the winner instead, even if actor X chose to defect from cooperation since yAx ! xEy would give i2, j3. In this game, the sucker is actually the winner of the game. This is the paradox which has been referred to. Depending on the rank order, which goals are preferred to other ones: G1 > G2 > G3 > Gn (here G refers to the `goodness' or `desirability' of the goal)? G1 is national security, G2 is welfare and G3 national wealth. Although the security and welfare of citizens are closely connected, the ordering above rests on the premise that the causal ¯ow seems to be more the former than in the reverse direction. Short-term objectives and longterm goals, national and local (or international) are often, if not totally incompatible, not capable of solution. This greatly complicates the practical necessity of deciding which goal is to be pursued and which ones are to be given a lower priority or even sacri®ced (Olson and Lee, 1994). G1 and G2 are sacri®ced last and states allocate large amounts of resources to enhancing security and improving welfare. Often rationality depends on the actor's preference ordering; some actors see certain goals as more important than other actors, because of historical events or cultural di€erences. Di€ering norms, the idiosyncratic traits of decision makers, age (c.f. Heberle's old `generation hypothesis') and belief systems (or gender, as many observers and researchers have argued) affect this sort of ordering (Cohen, 1991; Lewis, 1996). It can be said that the human being is an animal that has now moved out of ecological balance with its environment. Humankind has been and con85

Kuismin

tinues to be a wasteful killer and a despoiler of other life on the planet. This common, normal and apparently acceptable behaviour has been licensed by a belief that human use of the Earth's resources is God-given, as has been stated in the Old Testament and encouraged by a capitalist economic system that emphasizes short-term pro®t as a bene®t. Paradox is not based on wishful thinking or on a restructured view of human nature. It is a matter of enlightened self-interest. Finally, this paper quotes an old and brilliant American Indian proverb, which is also the key slogan of earth nationalism: `We have not inherited the earth from our parents, but borrowed it from our children'. Bearing this in mind, it is possible to formulate rational policies enabling sustainable development. Impatience along with short-term planning have also been and will be besetting sins of Western societies. The human race must learn to put a real cost on the resources it consumes and the wastes it produces. References Aggarwal, V.K. and Allan, P. (1992) Cold War end games. In The end of the Cold War. Evaluating theories of international relations (P. Allan and K. Goldmann, eds) pp. 27±8 and 34±5. Dordrecht: Martinus Nijho€. Brams, S.J. (1985) Superpower Games. Applying Game Theory to Superpower Con¯ict. New Haven: Yale University Press. Bueno de Mesquita, B. and Lalman, D. (1992) War and Reason. Domestic and International Imperatives. New Haven: Yale University Press. Buzan, B. (1991) People, State and Fear. An Agenda for International Security Studies in the Post-Cold War Era. New York: Harvester Wheatsheaf. Buzan, B., Jones, C. and Little, R. (eds) (1993) The Logic of Anarchy. Neorealism to Structural Realism. New York: Columbia University Press. Caldwell, L.K. (1990) International Environmental Policy. Emergence and Dimensions, 2nd edn. Durham: Duke University Press.

86

Camilleri, J.A. and Falk, K. (1992) The End of Sovereignty? The Politics of a Shrinking and Fragmenting World. Aldershot: Edward Elgar. Cohen, R. (1991) Negotiating Across Cultures. Communication Obstacles in International Diplomacy. Washington: United States Institute of Peace Press. Dougherty, J.E. and Pfaltzgra€, R.L. (1981) Contending Theories of International Relations: A comprehensive Survey, 2nd edn. Cambridge: Harper & Row. Elster, J (ed) (1986) Rational Choice. Oxford: Basil Blackwell. Gabriel, M. (1994) World Views and Theories of International Relations. New York: St Martin's. Glassner, M.I. (1996) Political Geography, 2nd edn. New York: Wiley. Greenpeace (1996) Annual Report. Hurrel, A. (1995) International political theory and the global environment. In International relations theory today. (K. Booth and S. Smith, eds) p. 149. Cambridge: Polity Press. Kennedy, P. (1993) Preparing for the Twenty-First Century. London: Harper Collins. Keohane, R.O. and Nye, J.S. (1989) Power and Interdependence, 2nd edn. Glenview: Scott, Foresman. Lewis, R.D. (1996) Kulttuurikolareita. Helsinki: Otava. Luard, E. (1990) The Globalization of Politics. The Changed Focus of Political Action in the Modern World. London: MacMillan. Martin, J.M. and Romano, A.T. (1992) Multinational Crime. Terrorism, Espionage, Drug & Arms Tracking. Newbury Park: Sage. Mathews, J.T. (1989) Rede®ning security. Foreign A€airs 68(2), 171±7. Nicholson, M. (1992) Rationality and the Analysis of International Con¯ict. Cambridge: Cambridge University Press. Olson, W.C. and Lee, J.R. (eds) (1994) Theory and Practice of International Relations, 9th edn. New Jersey: PrenticeHall International. Porter, M.E. (1990) The Competitive Advantage of Nations. London: MacMillan. To‚er, A. and To‚er, H. (1994) War and Anti-war. Survival at the Dawn of the 21st Century. London: Warner. Tomasek, R.D. (1987) The international drug issues as an illustration of complex interdependency theory. Int. Third World Studies J. Rev. 2(1), 15±16 and 28.

The Environmentalist