Published in: Hans Keman (ed.), 1993: Comparative Politics. New Directions in Theory and Method. VU University Press, Amsterdam,  p. 101-119

© Universität Osnabrück - Staat und Innenpolitik

>>>Digitales Archiv<<<


Institutional Difference, Concepts of Actors, and the Rationality of Politics

Roland Czada

Because man is the meaning-seeking animal,
he gravitates to loci (or institutions) where
meaning is likely to be found or created'
(from: Alan Sica, 1988: 264)



 The rational choice approach of decision-making and strategic interaction is intended to apply everywhere and at any time.  The meaning and solutions of, for instance, a prisoner's dilemma game or the public goods-problem do not vary across countries (Ostrom, 1990).  Being pure theoretical constructs, they are independent from the bounds of time and culture.  So far, the rational choice paradigm follows a natural science model of explanation. lt assumes man to be a logically calculating animal.
     If all men chose rationally the comparative method would loose its significance, for decision-making processes would be explained by a general algorithm of perfectly calculated choices. Cross-national differences would then be due to the free parameters of this algorithm, e.g. actors' preferences, available means etc. They could be explained in every case without cross-national comparisons. Yet it seems that the rational choice paradigm actuates major advances in the field of comparative politics, rather than eradicating this approach within political science.
     The question is therefore whether the field of comparative politics could profit from the rational choice approach without being sacrificed to it.  In this Chapter I shall demonstrate that the functional and cognitive limitations of rationality are an important source of cross-national variation and that, with some theoretical modifications, the rational choice paradigm can enhance our understanding and use of the comparative method in political science. I shall elaborate upon these points, starting with a critique of the rational choice approach.

5.1     Political Choice:  Calculation with Infinite Solutions

Rational choice is a function of the strength and direction of an actor's preferences, available means and individual assumptions on the cause-end-effect relationships of action.  From this perspective, cross-national or intertemporal variations of politics and policies can be seen to be caused by differences in the preferences, means and causal concepts of political actors.  For example, during the 1970s the British Government tried to win the trade union's support by proposing a social contract in order to combat inflation.  The government failed, because the unions preferred high wages and lacked the organizational means to control their membership.  In addition, considering associational and governmental control deficits most of the union leaders did not really believe that wage restraint would cause less inflation and more employment.  In contrast, the highly organized German unions preferred not only high wages, but also co-determination, and they believed firmly in the concept of Keynesianism.  During several years they were able to lower the wage-demands of their members in return for co-determination and expansionist economic policies (Scharpf, 1987).
Politics is a process based on relations of conflict and consensus among interdependent individuals and corporate actors.  Therefore, political choices are always concerned with interaction (Keman, 1992a).  Treating unions as rational corporate actors, their own interests, organizational means and policy-concepts will not fully determine their choices.  Since overall policy-outcomes also depend on the choices of their counter-actors, unions must calculate the interests, means and policy-concepts of employer associations and governments; and this again includes an assumption about how these other actors would estimate the interests, means and policy-concepts of the unions which they themselves are also trying to calculate at the same time.  In this way, every next step of calculation increases the time-horizon and number of possible solutions of a choice-problem.  Actors, in attempting to calculate their decisions, meet with cognitive and functional limits of rationality quite quickly, because of their interdependent relations.
     Whereas preferences can be treated as exogenously given, and available means are relatively easy to calculate, the idea of how a certain choice effects the realization of a desired end appears to be highly speculative, because it depends not only on the adequateness of a policy-concept but also on how other actors will react to one's own action.  This not only causes an 'exploding complexity of simultaneous optimization' (Scharpf 1991: 278), but can also result in a circular argument as far as reciprocal assumptions of the choice calculations of co-actors are concerned.
     The attempt to choose perfectly rationally would leave an actor with endless calculations to perform.  Beyond a certain level, this would be highly inefficient because of the inevitable rising costs of decision-making.  This, however, theoretically means that human action is incalculable.  Nevertheless actors do want to give their choices specific meaning.  This is a procedure of rational (self-) interpretation.  Actors do also assume that there is a meaning in the choices of other actors and, hence, interpret these choices accordingly.  This interpretation depends on, and adds new elements to, their own world views or ‘cognitive maps’ (Axelrod, 1976).  This causes a significant research problem for the analysis of actors in comparative politics.
Whereas structural relations between actors can be objectively measured, and functional correlations between variables can be made across different units of analysis, actions have to be interpreted in their specific contexts. lt is impossible, for instance, to explain the motives of a Finnish president dealing with a party in parliament in comparison to the American political system and the notion of political rationality that it carries.  Strictly speaking, researchers have opted for the 'policy-maps' (Schneider 1988: 82) of actors and for the institutional framework of action.  In his research on chemical control policies in a transnational policy-network, Schneider (1988) investigated the views of all relevant actors on the 'problem' at hand.  He found that conflicting parties can share views on the nature of the problem; although they differ with regard to their various interests, policy goals, means, and institutional incentives.
     The rationality of political actors is a subjectively disclosed and apprehended quality of action and thought (Goodwin, 1976: lOff.), although it is molded by social membership and institutional incentives. lf attempts to calculate choices turn out to be nothing more than speculations on an uncertain future, actors can rely on their intuition or they can develop, learn, or adhere to collective world views.  The latter is, of course, a cultural-institutionalist approach to the problem of rationality which is opposed to economic explanations based on the assumption of pure wealth maximizing actors (see Chapter 4 in this book).  These actors should stop calculating their choices only to save information costs.  Thus, economic man would calculate how long it pays to calculate.  This, however, is not very convincing, because one cannot rationally decide where to stop the search for information.  As long as every future search can possibly add decisive new clues about how to decide more rationally, it is necessarily a more or less nonrational decision to stop searching (Elster, 1986).  Hence, rational choices can lead to a vicious circle of reasoning without a final result other than making a subjective decision in the end.
     To sum up the principle argument on the status of rationality in politics and social life: problems with information and interactional dilemmas between conflicting parties result in uncertainty and ambiguity.  Ego's action becomes meaningless, if he or she lacks any reliable knowledge about its consequences in terms of alter's reaction.  Due to these obstacles, institutions have emerged as loci where meaning is likely to be found or created (Sica, 1988: 264).  From this perspective, institutions are a prerequisite of rational, meaningful action.  There is, and cannot be, any rationality without them.  For instance, in some Indian tribes it is rational to give away all of one's belongings in order to win social status.  The 'potlatch' as the procedure is called, makes entirely rational what Europeans would undoubtedly call irrational.  Thus, the rationality of an action has something to do with the meaning which institutions attribute to that action, be it through informal cultural features or by means of formal organizational rules.  This is particularly important for cross-national comparisons, because otherwise we might possibly call the deliberate choice, of, say, an Italian union leader, irrational, on the grounds of, say, the Dutch industrial relations system.
      Institutions determine the rules of the game and, thus, furnish the conditions of the choices of individuals as well as of corporate actors under those rules.  Institutions constrain choices and make actions more predictable.  They routinize action and enhance the opportunities of actors for consequential, strategic interaction.  Political institutions give rights and impose duties on individual and corporate actors.  Simultaneously, they must be regarded as arenas for conflict, political leadership, ideology and goal-setting (Strom, 1990; Keman, 1992a).  As a structural framework for strategic choices, institutions serve as opportunity structures rather than as frictionless decision-making machines.

5.2              Cognitive Concepts and Strategic Choices

Institutions determine the capacity of political actors to act and interact with each other.  Thus, they do not fully determine political action, but instead leave room for strategic choices.  Therefore, a singular knowledge of institutional rules is insufficient to analyze, explain or forecast political behavior.  Political goals and actors' ideas on cause-end-effect relationships are equally important.  For instance, if political parties or labor unions believe in Keynesianism they will promote deficit spending and high wage policies in order to overcome mass unemployment.  In such cases, institutions are to be seen as enabling or restricting certain policies (Scharpf, 1987; Braun, 1989).  Policy-concepts may be rational in themselves, but can cause unforeseen, non-rational effects depending on institutional environments.  Moreover, policy-concepts may appear to be rather parochial, at least this depends on the perspective of an actual observer.  His or her primary task is, however, not to judge the rationality of concepts in themselves.  Political scientists should leave this to economists and should instead concentrate on policy-concepts as determinants for political choices.  This includes research on the relation between concept-building and institutions, as well as on the relation between institutions and the realization of concepts.
     So far we have distinguished the conceptual policy-rationality, based for instance on the Keynesian theory, from the political rationality of actors seeking to promote certain ends in institutional settings.  Comparative policy analysis has to scrutinize political institutions and policy-concepts (Keman, 1984; Schmidt, 1987).  They both influence political actors' choices.  It makes a difference, for instance, whether or not a central bank chooses to support high wage demands and governmental deficit spending by an expansionary monetary policy.  This choice depends on the institutional autonomy of a central bank, its political goals, and its adherence to the causal assumptions of the Keynesian concept.  One could say that the room for choice and strategic interaction is determined by an actor's institutional position, political goals and cognitive concept of action (Dean, 1984).
     Institutions, goals of an actor and concepts of action influence each other.  This makes it methodologically difficult to use them as explanatory variables in comparative research designs.  Scharpf (1987) and Schmidt (1987) suggest that it is the will and skill of politicians that determines political success.  In this respect we could say that political skill is the ability to construct and use political concepts that fit a given institutional opportunity structure.  Thus, institutions determine the extent to which concepts are in fact rational, i.e. they can help to realize certain goals in practice.  Rational political actors must combine theoretical concepts with practical constellations of interest intermediation, power-dependencies, scarce means and so forth.
Political actors can have a multitude of goals and concepts, which are meaningless in regard to the rational choice paradigm.  In fact, there is non-rationality in politics in the sense of actions not comprehending causality in attempts to secure ends.  Another type of non-rationality consists of a defiance of rational procedure.  In this case actors use causal concepts to realize their goals, but they do so without considering the institutions which can be supportive or hostile to the concepts.  To apply a political strategy of polarization in a consociational democracy, as the Dutch labor party did during the 1980s, would be a case in point (Braun, 1989). If this was due to a lack of institutional knowledge, it however provides us with a rational explanation of nonrational behavior.  Thus, the existence of non-rationality does not necessarily mark the limits of a rational choice approach of explanation in political science.
     Political actors can pursue joint interests and share cognitive concepts.  In this case, consensus about policies will be high and the actors easily coordinated.  Integrative institutions as, for example, corporatist arrangements do also support consensual policies.  The more actors, interests and concepts that are present in a decision-making process, the less likely it becomes that coordination is feasible as a decision-making style.  Politics then approaches a market-like pluralist process.  Policy outcomes become unforeseen aggregate effects of choices of single actors.  They emerge from the interplay and intersection of more or less isolated political actions.  The rationalities of a particular actor's choice loose significance if their overall consequences cannot be calculated but follow from the configurative rationality of a whole social system instead.  In this case, institutions become rather important, not for the explanation of particular strategic choices on the micro-level of action, but of the formation of public choices.  Many political institutions have not been designed to influence individual choices, but instead to translate the wills of individual actors into collectively binding decisions to be executed by a corporate actor, be it the state, interest associations or any other corporate body.  Electoral systems would be a particular case in point.  Such systems are explicitly designed not to predetermine the result of elections.  All modern institutions of democracy, public administration, and the rule of law are intended to register and aggregate rather than to determine individual choices (Czada/Lehmbruch, 1990).
     The rational choice paradigm embraces both, the logic of interaction between a few actors and the logic of public choices, i.e. the transitory coupling of atomized actors or the joint action of groups and organizations.  The first is dealt with by game theory, the second by approaches relating to the public goods problem (Arrow, 1951; Downs, 1957; Olson, 1965; see also Chapter 4 in this book).  In the next section, I will focus on the empirical complexity of political interaction and I shall discuss some systematic, macro-political aspects of choice in relation to comparative politics.

5.3             New Challenges in Comparative Politics

The study of comparative politics is today confronted with more and more differentiated structures of governance.  In most of the liberal democratic countries, majoritarian national governments are just one source of legitimate authority (Lijphart, 1984).  Sub-national governments, supra-national regimes and governmental bodies, national and international associations, big multinational firms can be seen more or less interwoven policy-makers determining overall policy outcomes.  This complexity makes policy-analysis rather difficult (see also Chapter 7 in this book).
In areas where the nation-state becomes increasingly replaced by supra-national authorities, the institutional framework of politics still includes national and sub-national elements.  In fact, new political institutions often cross old-established borderlines of territory or functional domains.  For instance, nuclear melt-downs at Three Mile Island in 1979 and Chernobyl in 1986 caused farreaching reforms of the institutions and measures of nuclear safety regulations.  Among others, a private institution, the 'World Association of Nuclear Operators' was established in 1987, 130 firms operating nuclear plants in 29 countries participated.  If one compares the politics of nuclear regulation cross-nationally, one has to investigate the role of such an international actor.  At the same time, German sub-national state governments led by Social Democrats and the Green Party tried to use nuclear safety regulations to phase out of nuclear energy, whereas the national government defended the established policy-concept.  In sum, a complex multi-level network of actors with moveable borderlines and without a clear center of decision-making is emerging.  Political choices have become embedded into a bargaining system, which cannot be deliberately changed by any one of the participating actors (Czada, 1992), but is in itself a bargained system.
The move towards supra-national and sub-national arenas of decision-making, certainly does not reduce institutional varieties.  On the contrary, the number of institutional arenas increases.  The history of politics is a history of institutional differentiation moving towards more and more specialized, and at the same time interwoven, organizational units of decision-making.  This challenges the discipline of comparative politics in several respects:

Below, I shall demonstrate how these points relate to the conceptual history of comparative politics.  In addition, two approaches will be introduced, which seem particularly promising in regard to an institutional explanation of policy-making:

1. The notion of moveable boundaries between units of action and overlapping memberships of actors in bargaining networks, as well as interdependencies between institutional domains, should add new perspectives to comparative analysis.  Such a perspective is by and large compatible with the network approach, discussed by Adrienne Windhoff-Héritier in Chapter 7 in this book.

2. Cultural norms of behavior play an important role in explaining cross-national variations in policy-making.  Even in international bodies with unitary organizational structures and procedural rules, members of different nationalities behave differently in identical situations (Hofstede, 1980).  For instance, American nuclear inspectors have applied identical regulatory standards as those of their Gennan colleagues for a completely different task (Czada, 1992), and it is doubtful that this is only an outcome of different organizational structures (Kelman, 1981).  Often the norms of behavior embrace the centuries of historical experience, which political or social communities have endured.  Hence, it requires historical knowledge to explain and understand actual developments.  This is an old theme mastered by the classicists of comparative analysis - for example, de Tocqueville, Weber, Lipset and Rokkan.

5.4               The Antiformalist Revolution and its Consequences

The formal principles of government and public policy-making are in most cases laid down in constitutions and administrative law.  So far, the process of government has been influenced by legal rules, which vary considerably across countries.  Constitutions give rights to individuals, they constitute individual freedom and regulate how individuals can legally participate in public affairs.  Conversely, public administrations could be called the 'operating state', executing decisions made according to constitutional rules.  Political scientists have for some time studied those formal mechanisms under the heading of 'Comparative Government' (Finer, 1970).  After the Second World War, an alternative antiformalist approach came to the fore. It focussed on social facts of interest intermediation and political regulation.  From this perspective, the process of government is driven by interest groups trying to influence public policy without being directly involved in governing.  The so-called pluralist movement 'was a sociological revolt against legal formalism: group interaction constituted the reality of political life, operating behind the formal legal-institutional disguises of society and the state' (Almond, 1983: 173).  A bulk of interest group studies appeared covering many countries on all continents.  Suzanne Berger (1981: 18) alleged, for instance, that in 'formulating social demands and channeling them into the political process' interest groups gave definition to the varieties of political processes and its outcomes in different countries.
     Whereas former studies of comparative government could rely on normative theories of law, the separation of powers, representative government or of legal bureaucracy, the antiformalist movement was essentially empirical in its approach.  The historical richness of, for example, the works of Truman, Beer, Ehrmann, Latham, Leiserson is still impressive, but leaves behind a sort of discontent as far as the development of theory is concerned (see also Chapter 2 in this book).  This feeling of discontent has been expressed by Mancur Olson quite strongly in his book on the 'Logic of Collective Action' published in 1965.  He claimed that the pluralist 'group school' started from wrong assumptions in regard to the associability of individuals and resulting mechanisms of group action.
      Olson's argument is that sharing an interest will not automatically lead to collective action in favor of its achievement.  From a rational choice perspective, the marginal contribution of an individual to a lobby group would hardly increase its chances of success.  Furthermore, it is not feasible to exclude non-members from the consumption of public goods, and certain overall policy outcomes, which result from pressure politics.  Therefore rational individuals tend to avoid becoming due-paying members in interest associations.  They instead prefer a free ride on the actions of others.  Such behavior is less likely, if not impossible, in smaller groups.  The mutuality and social control amongst their members favor individual involvement.  Most small groups often pursue specific goods whose consumption is exclusive.  Hence, it is easier to organize as well as to manage small narrow interests groups.  In contrast, large groups are difficult to organize due to the free rider problem.  Once established, they tend to be internally divided and therefore hampered in their attempts to represent interests in a coherent manner.
     The pluralist 'group school' did not consider that interest politics is molded by individual choices, which are in turn channelled by organizations and interorganizational networks.  The latter translate individual choices into public ones according to the rules that are laid down in national institutions of interest intermediation.  In this way, cross-national variations can be attributed to certain patterns of cultural choice and to incentive structures developed in national institutional settings.  The institutionalist and rational choice perspectives clash with the pluralist approach.  Essentially, the latter approach subscribed to a functionalist view of politics.  'It presented the notion of multifunctionality as a property of all political structures' (Almond 1983: 180).  Regardless of their structural characteristics, competing groups were seen as demanding and supporting public policies in a functionally equivalent manner.  Thus, even under authoritarian regimes and in developing countries, informal groups and various forms of pluralist pressure politics appeared to drive the political process (Linz/Stepan, 1978).  Up to the early seventies, cross-national comparisons were dominated by this more or less ethnocentric view, which treated political systems as a variety of the American pressure group lobbying.  Implicit to this analysis was a functionalistic optimism, which attributed policy outcomes to the beneficial consequences of the pluralist process.
     It was not until the challenge of 'corporatism' in the mid-seventies that institutional characteristics of national systems of interest-intermediation came to the fore.  Such systems have now been defined as organized, hierarchically structured and more or less biased with respect to different wants in a society, i.e. interests are not equally represented in the system.  In several countries associations not only represent their members to governments, but also the government to their members.  Thus the corporatist 'school' puts interest groups into an intermediary position between the government and individual citizens (Streeck, 1984).  The pluralist emphasis on pressure and influence was replaced by the idea of interest groups being involved in governing (Katzenstein, 1985).  Thus, associations were considered to be a specific form of governance besides markets, states, communities or clans and firms (Streeck/Schmitter, 1988). Both the corporatist school of interest intermediation and Olson's logic of collective action seriously undermined the pluralist explanation of politics.  But in a way, the corporatist view was even less consistent with Olson's theoretical findings than the pluralist one (Olson, 1986).  Why should one pay for membership of an association that occasionally represents the government against ones own individual interests?  Nevertheless, interest associations in Sweden, Austria, Switzerland, the Netherlands and Germany often act as quasi-governmental bodies regulating certain aspects of public life.  Their incorporation into politics is part of a historically inherited institutional structure that cannot be deliberately changed by individual actors (Van Waarden, 1992).  From a universal rational choice perspective, it seems that they are entrapped by institutions forcing them into membership relations against their own individual interests.  However, an instutionalist-cum-rational choice perspective suggests specific cultural expectations of how to account for such pattern of interdependency.  Such an explanation focusses on decision-making styles like confrontation bargaining, problem solving (Scharpf, 1991) which are being produced and stabilized by processes of political socialization; a case of social learning within institutions which is determined by institutional content or spirit, rather than by institutional structures.
     To illustrate this point, the problem resembles the question of whether the roman-catholic church survived so successfully during more than a millennium because of their hierarchical organization, or because of the spiritual orientation and loyalty of their members.  The same problem applies to the comparative analysis of labor movements and trade-union organizations.  To concentrate on organizational properties, as researchers do on corporatism (e.g. Schmitter, 1981), ignores such problems.  If it is true that culturally inherited orientations may substitute the effects of formal organizations in controlling the opportunism of actors and directing their transactional choices (North, 1981), then comparative research has to deal with the complex interrelation of socialized actors and historically grown institutions.
     It is difficult to explain the emergence and stability of corporatist systems as a result of individual rational choices.  Methodological individualism, as expressed in rational choice theories, appears to be incompatible with an institutionalist explanation of politics.  Institutional macro-structures clash with individual micro-motives.  From the viewpoint of normative and functionalist theories this is not surprising at all, because institutions should serve as barriers for opportunism.  Institutions are meant to give meaning to a specific action and thereby are meant to reduce the possibility of alternative paths of action.  They constrain individual choices.  However, in some countries the institutions of interest intermediation constrain such choices more than in other countries.  There exists considerable cross-national variation in associational density and the degree of organizational regulation of capitalism, which cannot be explained by functional necessity.  Economies can be governed predominantly by corporate 'clans' as is the case in Japan (Ouchi, 1981), by voluntary associations as in Switzerland (Farago, 1987), or by bureaucrats as in France (Dogan, 1988; Zysman, 1984).  These forms of governance are, however, not in constant flux.  Their key feature is a pervasive stability.  Therefore we should examine patterns of political action within institutional systems from a comparative perspective.  This would allow us to understand the stability, causal links and operating mechanisms of governance and interest intermediation in various countries, however different they may look superficially.

5.5             Rationality, Culture, and Patterns of Interaction

Both comparative politics and policy analysis are intended to explain difference.  Researchers examine variations and investigate how they come about.  As has been demonstrated above, the rational choice approach holds some resolute assumptions about the sources of variation across countries.
     To assume that the political process is based on rational choices on the micro-level of societies leaves three possible explanations of the cross-national variations of that process.  Firstly, as preferences, means and cognitive concepts of political actors differ, the process and outcomes of policy-making will also vary.  Differences will exist between nations, if preferences, means and concepts of actors are linked to social and cultural environments.  Secondly, the rationality of choices is shaped or even determined by institutions.  Differences between nations would then simply reflect differences in their institutional design.  Thirdly, differences between nations can follow from specific, culturally defined understandings of rationality.  In this case, even identical preferences, means, cause-end effect concepts and institutions could produce different choices.  These possible sources of variation do not necessarily exclude each other.
     Critics of the rational choice paradigm have always emphasized the cultural roots of the concept (Hirschman, 1977; 1982; March/Olsen, 1989; Douglas, 1986).  The notion of a sovereign individual is the historical product of rationalized societies.  In this view, rational choices require a specific cognitive base that has to be learned.  Historical learning leads to the assumption that the shaping of rationalization differs across nations.
     Cultures vary in the extent to which action is expected to be carried by individual actors.  This approach points to a normative framework of rules and obligations which is external to individuality.  It determines the actor's understanding of what to do in any given situation.  According to this approach, cognitive concepts of how politics works, play an important part in the explanation of cross-national differences of political action. If, for instance, actors believe that a freely elected national parliament is only an arena for discussion, but which hardly ever functions as a decision-making arena, as has been held true for long by left wing and right wing politicians in Germany, dictatorship appears to be just another form of normalcy.  The Weimar constitution at its time the most modern one constructed by notable experts of law, history and the social sciences (Bolaffi, 1989; Luthardt, 1990), did obviously not provide the democratic institutions appropriate for political and administrative actors who had been socialized and used to an authoritarian monarchy; especially since this was an early developed welfare state and, thus, enjoyed a somewhat legitimate status.  To establish a highly proportional system of parliamentary representation, and supplement this with elements of direct democracy, was probably a wrong incentive structure for most German political actors at that time.  Otto Kirchheimer (1932) said the constitution limped ahead of its time.  Others asserted that it set up a democracy without democrats.
     This makes clear, what is meant by cultural attitudes of actors: these are group specific orientations which can exist independent of institutions, particularly in times of institutional change or break-down.  We then experience a cultural lag between the cognitive orientations of actors bound to past institutions and to the newly established institutional incentive structures.  Such lags, of course, mark the limits of institutional reform.  The argument, however, is not be understood to say that there is a somewhat genetically inherited spiritual orientation of nationalities, clans, families or any other groupings.  Instead, subjective orientations being individually internalized although collectively shared images of reality do not necessarily cope with changing institutional environments without ruptures.  This has been empirically shown in numerous developmental programs throughout the Third World (Prechtel/Harland 1986).  One can currently observe this, by examining the way in which West German institutions are being transferred to the former East German territory. It is said that the Berlin wall was much easier to remove than the wall in the brain, which blocks the unification-process more than any other aspect.
     In contrast to the view of new institutional economics (e. g. Williamson, 1975; North, 1981), the cognitive aspect is important, not just because it influences individual calculations, but even more so because preferences and the perception of a situation are already shaped by culture.  For example, social equality is highly valued in the Scandinavian countries, whereas Americans prefer individual achievement (Lipset, 1963).  Hence, the 'rationalities and irrationalities of the nordic welfare state' (Andersen, 1988) can certainly be understood as a result of cultural attitudes; and the same is true for American politics.  Hofstede (1981) found that 'uncertainty avoidance' varies along national cultures.  So one can assume that cultural belonging determine one's choices, particularly when high risks are involved.  This leads to another factor in the explanation of cross-national variations of policy-making.
     The rational choice paradigm is based on the assumption that actors have perfect information and are able to make decisions that maximize some concrete ends.  It is a common-place idea, since Herbert Simon's research on organizational decision-making showed that actors search for satisfying instead of optimal solutions because of informational restrictions (Simon, 1957).  On the one hand, this adds to a cultural-institutional explanation, since cultural rules and organizational routines reduce the uncertainty and ambiguity of individual choices.  On the other hand, the concept of bounded rationality leaves room for random walks in decision-making.  If the consequences of a choice cannot be calculated because of poor factual knowledge and ambiguities about preferences, then problems and solutions become not only more complex but to some extent reciprocal or circular.  Actors may start from existing solutions and apply them to unknown problems.  They may alter or even discover their motives by acting.  In this way, decisions can emerge from the random matching of solutions and problems.  The 'garbage can model' (Cohen/March/Olsen, 1972) of organizational decision-making suggests that in the face of new problems, when actors are highly uncertain, a recombination of existing solutions will often occur.  Randomness comes in, because:  

'the model assumes that problems, solutions, decision-makers, and choice opportunities are independent, exogenous streams flowing through a system.  Solutions are linked to problems primarily by their simultaneity, relatively few problems are solved, and choices are made for the most part either before any problems are connected to them, or after the problems have abandoned one choice to associate themselves with another' (Olsen, 1991: 92).

Such 'garbage cans' of policy-making do establish path-dependencies, although there is a kind of openness in non-routinized situations which gives room for deviations. It follows from institutional inertia and from the bounded rationality of decision-makers that certain solutions to political problems turn out to be dominant over time in certain countries.  In Germany, for instance, social welfare problems tend to be solved by compulsory insurance schemes.  In Scandinavia, or the Netherlands, tax-based systems with flat-rate welfare provisions is the selected policy strategy.  Only crisis situations, in which complex matchings of actors, problems and solutions occur, give room for new directions in policy-making.  Transformations of national policy-styles can, for instance, occur in the wake of mass-unemployment, hyper-inflation, natural or technological disasters or civil unrest.  Federalist systems can become centralized or even deferred into authoritarianism during economic depressions as happened in Nazi-Germany.  Analyses of political crises show that it proves difficult to regain control of social perplexities' by reverting to established rules of policy-making (Czada, 1992b), or, if they do, it leads to diminishing returns in terms of cost-effective decision-making.  The discipline of comparative politics, however, lacks systematic studies of institutional break down.  The loss of rules, which characterize such situations, extend the strategic options of political actors.
    Besides its cognitive-cultural aspects and 'garbage can' characteristics, policies are developed in formal organizations and inter-organizational networks as will be further elaborated in Chapter 7 in this book.  Research on corporatist interest-intermediation has shown how organizational structures and interorganizational networks influence the process of governmental decision-making and its policy-outcomes (Lehmbruch/Schmitter, 1982).  Lehmbruch (1984) reports on a 'neocorporatist logic of exchange' between conflicting but functionally interdependent and institutionally linked interest organizations.  This gives way to the idea that the relations between political actors are not solely dominated by zero-sum distributional conflicts.  There are also complementary and parallel interests (Scharpf, 1984).  Political actors may, in fact, face each other as potential competitors, as exchange-partners, or as associates (Czada 1991: 263).  In reality, the recourse to these patterns of behavior is open to strategic choice.  On can find them in different mixtures, depending on actors' goals, available resources, changing environments and power-dependencies.
The debate on corporatism shows that the association of convergent or joint interests appears to be much more a theoretical problem than a practical one for explaining relationships of political exchange or competition.  Exchanging or competing for political resources like votes, influence, status, power leads to market-like, aggregative processes.  In contrast, interest associations supersede the transitory coupling of individual actors.  Political authority and an executive body are required in order to maintain and manage an association.  Every association of interests requires some kind of governance in order to realize its goals, be it the state or other organizations.
     To view politics as an integrative, institutionally shaped process instead of a mere aggregative mechanism has several implications.  Institutions have to be seen as autonomous forces.  They are public goods, which single actors cannot deliberately create or change for their own convenience.  In searching for the rational foundations of political action, one may assume that institutional patterns of interest intermediation offer specific pay-offs for individual and corporate actors that vary from country to country as well as from one sector of policy-making to another.  The advantage of viewing systems of interest intermediation as incentive structures (i.e. delineating pay-offs) instead of systemic regulative mechanisms (i.e. adapting to problems) is threefold:

1. The conditions of political stability can be better understood by viewing the motives and choices of specific actors instead of functional adaptive mechanisms.  We have seen, for instance, that industrial relations systems do not adapt automatically to specific economic problems, but unions perceive those problems according to the background of their historical experience and ideological concepts.  The choices are influenced by available means and more or less speculative assumptions on the choices of other actors in society and politics.  Historically persistent patterns of political choice behavior are due to the rationally motivated resistance of political actors to change.  The rational motive is twofold: change entails uncertainty and high risks with respect to newly emerging social constellations, and it is expensive with respect to moving-costs and the economic as well as the cognitive burdens that are associated with change (Hechter, 1987).

2. Differences between processes of interest intermediation are no longer seen as deviations from a functional ideal-type like the pluralist equilibrium, but are instead viewed as the result of institutional incentives and consequential political action.  There are different patterns of interaction like pluralist, corporatist, or sectoralist intermediation with specific rationalities and decision-making-styles.  The latter are influenced by cultural orientations.

3. Forecasts and interventions can be made, if one knows how incentive structures and behavioral patterns determine actors' choices, and how these are then translated into public policies by institutional processes of interest intermediation or interest aggregation.  This, however, is an ambitious claim.  Whereas institutional incentive structures and behavioral patterns of actors can be observed and compared with the aid of conventional techniques of empirical research, it is much more difficult to close the micro-macro link between the choices of actors and public policies resulting from the effects of interaction between those choices.  One could, for instance, forecast, that the German unions will discipline the wage-demands of their members in favor of the economic closing-up of the East. It is, however, difficult to say, how important wage-rates are as a determinant for economic growth in the east; and one could only speculate how, for instance, the central bank, will react upon union policies.  The bank could ease monetary policy and, thus, support economic growth. It could also take a restrictive course to counter inflationary tendencies originating in global financial markets or international crises, which develop quite independently from German union policies.  Thus, one at best can only construct scenarios.
     In any case, the prognostic power of an actor's approach appears to be much better than that of pure structuralist or functionalist approaches to politics.  The inquiry of the political process should also take into account the cultural and institutional development of the political systems under investigation.  Hence, the degree of rationality cannot be detected by a pure, or economic, concept of rational choice, but must instead be analyzed and interpreted on the basis of 'real life' experience.  One way to do this is by developing rational-cum-institutional concepts that can 'travel' cross-nationally.  Perhaps this seems a daunting task, but it is one that ought to be undertaken.