Theoretical Aspects of Non-Lineal Models and Application
in Public Policy Analysis
Giovanni E. Reyes
University of Pittsburgh, Graduate School of Public and International Affairs
Contents
Introduction; First part: Main characteristics of non-lineal methods -chaos theory; theory of systems; theory of scenarios; comparisons and contrasts-; Second part: Applications to economic adjustment measures in Central America.
Introduction
Non-lineal methods of analysis have the advantage to take into account quantitative methods of analysis but they are not limited by numerical expressions only. In addition, non-lineal methods can express in a global sense, causal factors as well as results with time orientation and taking into consideration limits that we, in a previous way, have established as "border" on our units of reference. Chaos theory, as an example of non-lineal methods, from its basis in multiple equation systems, and from its biological perspective, can be used in studying social and economical problems. Complex systems are entities that can analyze using non-lineal methods.
Based on the aforementioned general considerations, this essay aims to cover two main objectives. First, to summarize the major characteristics of three non-lineal methods: chaos theory, theory of systems, and theory of actors and scenarios. Second, to apply those theoretical basis to a practical problem in public policy analysis. This problem will be the economic adjustment measures implemented in Central American countries. The principal objective of this second part, will be the application of non-lineal methodologies and to discuss two categories that I will like to introduce in my analysis: Levels of power and spheres of action. These categories are going to be explained in the second part of the document.
The main argument of this document is that the economic adjustment measures carried out in Central American countries are good examples of complex systems, they can be studied by non-lineal methods, these measures are public policy decisions which affect societies as a whole with significant and permanent effect. This application of non-lineal methods to the economic adjustment methods do not limit our analysis only to qualitative considerations; it can include numerical studies as part of the elements of criteria to be used in the analysis. The principal intention of this document, however, is not to illustrate this inclusion of numerical treatments of data, but to give a brief example of practical application of non-lineal methodologies to public policy analysis.
First part: Main Characteristics of Non-Lineal Methods
In order to summarize those theories, I will take as main assumptions the social conception of the economic adjustment plans in three complementary aspects. First the sense that the economic adjustment plans are affecting the society in each country as a whole with direct an indirect but no less significant effects on several social groups, thus we need to keep in mind the global sense of the economic impacts, and the level of social integration in which those impacts are working, and their changes over time.
Second, the fact that social conditions are not completely subjects to be quantifiable as a whole set of variables with respect to their total results (direct impacts), and repercussions or collateral effects (indirect impacts). This condition can be solved by analyzing the whole set of evident factors as a whole using integrated social theories in the non lineal systems.
Third, the social forces behind the economic and political effects are the main social groups, the most representative political actors and the historical trends that social movements have within the conditions of each society.
The use of the non-lineal methods to study social problems include the aforementioned aspects, but essentially it is important to note that these characteristics reinforce the global sense of the perception of the events under analysis, that is to say, the integral perceptions of factors, and results. In a complementary way, the main assumptions attempt to define the major motivations and driving forces behind social and economic actors or agents. Through a complementary perspective, the non lineal methods try to determine the more probable future scenarios, for this determination the theory leads us to base our perspective on the current conditions, their past trends, and the expected influence of the main factors in the future conditions.
One of the most important advantages of studying the social phenomena through the non-lineal models is that it allows one to formulate with more openness and creativity a greater number of hypotheses. To carry out the research process we can include here the inductive, deductive and the analogical ways to establish previous explanations of phenomena.
The general structure for the theoretical exposition of this complementary paper will be as follows: first, I am going to identify for each theory its main features referring it to its methods, objectives and its application. Then, I will summarize its principal weakness and strengths. As a third stage, I will address some particular applications of the theory under study, in order to have, as a fourth stage, a comparison and contrast among the three theories presented here: chaos, systems, and actors-scenarios. The concrete application of these three methods will be developed in the second part for this essay.
Chaos Theory
The main characteristic of the chaos theory is its study of phenomena considering all the factors involved in its complex dynamic and structure, through this feature, this theory allow us to identify the general patterns and future behavior of social and economic phenomena. In order to carry out this identification process, one needs to consider the essential causal elements within the phenomena, the major relations among crucial elements, and how they work, and develop their dynamics. Chaos theory constitutes a dynamic scheme of exposition and interpretation of reality, by determining the general trends of causality, fundamental and collateral factors involved, and by studying them from a global, that is from an integral perspective.
Among the patterns of behavior of phenomena we have the "butterfly effect", which was discovered by Edward Lorenz in 1961 when he was using non-lineal equations to predict a wide set of meteorological variables. One of his most important conclusions related to the butterfly effect was, not only in terms of the identification of a pattern of behavior or the cyclical recurrent effect, but also his observation that even small changes in the most important elements within a phenomenon can have dramatic repercussions in future results. The butterfly effect shows how initial causes can twist and turn in public management systems and over time generate surprising effects.
One of the main advantages of the chaos theory and complexity is that it allow us to develop a vision in order to understand the global dimension of the phenomena. From this perspective, the chaos theory does not reject quantitative methods, conversely, this theory includes them, but as elements to establish an integral or holistic trend in function of time referring to the most influential factors.
Stuart Kauffman has developed aspects of the theory of chaos and complexity from a biological point of view. In his recent book "At Home in the Universe", Kauffman explains that chaos theory is the general theory based on the idea that highly complex seemingly random systems settle into orderly states. This author shows that complexity and chaos itself is the cause that if enough different elements pass a certain threshold of complexity, they begin to self organize into a new entity. Chaos theory uses the basic insight of "order of free", as Kauffman calls it, to address a wide range of phenomena, from Darwin’s theory of evolution, to the new frontier in genetics, and to the level of the world of economics and social and cultural studies.
In terms of the weakness of this theory it is possible to mention two aspects: a) Under specific conditions the level of generality of the theory, even when it can develop a good global forecast, cannot specify the concrete tendency of the major aspects and elements which in turn are responsible for the internal dynamics of the phenomena; and b) That this theory can allow a significant margin of "subjectivity" from the researcher’s point of view.
Examples for the application of the chaos theory are the aforementioned case for meteorological forecast, and also the economic and social events, in which some of the risks can have great margin of unpredictability. More concrete examples are: mortality and morbidity rate from the affection of a disease, within certain areas or regions; the direct and indirect impact from the implementation of a residential construction project in an urban or rural area; the impacts from the production on the distribution and consumption spheres for a particular good or product, when we have changes in the economic national policies which are affecting the production processes, such as credits, taxes, external tariffs or quotas; and the impacts from a social policy related to subsidized social services in health services, including the medicare system.
Theory of Systems
The theory of systems is placed in an intermediate level of specificity between the globally and the integral perspective from the chaos theory, to the particular and more concrete analysis from the actors and scenario foundations. One of the most important features of the systems theory in terms of its theoretical basis and applications is a new paradigm between the struturalism and post-struturalism for the great discourses or social paradigms, and to the unique and non repetitive features from the ethnographic methods in cultural and social studies. The theory for systems articulates actors and the dynamic forces derived from social events to establish the major causes and the main internal factors for changes within the process of a phenomena.
Theory of systems is an attempt to identify the main elements within a phenomena, how these elements are related to each other, and if it is possible, to establish some particular links among the main factors. When this last mentioned situation is achieved, we can define sub-groups or sub-systems from the internal conditions of the phenomena under study.
In addition to the concept of sub-systems, this theory involves the concept of levels of integration, which are basic foundations to the study of ecological, economical and social events. These levels of integration would be constituted by, for example, the individual, the family -nuclear and extended-, the social group, social class, state, nation, and a particular world region. We can study the different sub-systems and systems within each particular level of integration and from this basis predict the impacts of a proposed political and social policy, or those repercussions from an already established policy.
There are two main advantages of the systems theory: a) its intermediate level between the more global perspective of the chaos approach, and the more concrete vision of the actors-scenarios methods; b) the fact that theory of systems can gradually establish the perception of a phenomenon using different levels of integration. In general this theory, as the other non-lineal methods, allow us to study different elements of a phenomena trying to integrate rather than to separate its elements, and setting them in function of time, in both senses: to the past tendency, as well as the future perspective.
The main disadvantage and critique of the systems theory has been the consideration that its approach is basically a static one. Several attempts have been made to resolve this problem. The concept of momentums is one response. This notion contains the theoretical foundation that systems can be determined in several moments during the development of a particular event. In each one of these periods of time, or momentums, it is possible to study the current conformation of factors, relationships among them, sub-systems and levels of integration involved in particular systems under study. This methodology is complementary to the level of specificity that the levels of integration allow us to see inside a phenomena, and its useful to determine the more realistic perspective of development of events.
Considering the external and internal conditions, systems theory determines, also, the levels of systemic spheres -that is the exterior influence-, and the sub-systemic conditions -the internal circumstances for a particular event-. This division can be seen very clearly in the three main theories of international relations -liberalism, globalism and realistic and neo realistic approaches-. However this notion can be applied to other spheres of study such as economics, social and cultural issues. In these last cases, the "borders" of systemic and sub-systemic spheres are less clear and thus are more subjects to the personal or "subjective" interpretation of the researcher.
Examples for the application of systems theory are studies of social groups which have common economic, or political or cultural interests, and how they act in certain conditions of changes or stability within a society; how pressure groups react to specific policies in social or economic terms, or how different institutions are related to the political power taking into account the external and the internal conditions in a particular society.
Theory of Actors and Scenarios
Among the three non-lineal theories here exposed, theory of scenarios is the one which presents the most concrete approach to determining specific conditions and characteristics of actors and the scenarios they conform. The main characteristics of the theory of scenarios are related to the definition of the central axis along which several actors develop their major interests, connections, and general behavior. Their different degree of association allow them to conform stable or non stable sub-groups which in turn are the basis for the particular conformation of scenarios in a more global sense. These scenarios can have a more permanent situation or a less stable one. If the situation is more stable we can say that they conform a structural scenario, and if this conformation is a provisional one, we say that their relations are based on junctural connections in the scenario under study.
Based on actors’ interest, fundamental motivations, general behavior, relations, structural or junctural features of relationships, theory of scenarios allows us to determine a general forecast of future relations in a global sense. This situation results in the forecast of several scenarios based on specific criteria, and considering the influence of external and internal conditions of change. By specifying several scenarios we can take into consideration future reactions to a social policy or particular change in complex systems, and how actors can react changing or not changing their relationships, interests, and behavior.
Two of the major advantages of this theory are: a) This non-lineal model can be as specific as the researcher wants and/or have data -quantitative or qualitative-, in terms of determining the concrete conditions from actors and their relationships; b) As main axis of action: actors, relationships, and interest; the researcher can include as elements of criteria the presentation of quantitative data, as his/her basis for qualitative interpretation or results. One of the methods to quantify data from relationships is the model of network analysis.
Among the more cited disadvantages of this theory are: a) The specific conformation of actors, their behavior, and relationships are subject to significant influence from the "subjective" and "ideological" perspective of the researcher; and b) Because of the specificity of the actors and scenarios, the perspective of more structural and permanent trends within the development conditions of a particular phenomenon can be lost, especially those which belong to the social and economic field.
Examples for the application of the theory of scenarios include the social reaction to a change in a social security policy, a tax system, and a credit policy which is going to affect all the banking mechanisms in a country; how pressures groups are going to react to privatization of public institutions or enterprises, international agreements, or changes in the legal system. Theory of scenarios can be useful to determine what actions the main actors can take in anticipation of the reaction from several groups, in order to conform stable or desirable scenarios in the future.
Comparison and Contrast
In terms of level of analysis, the chaos theory has a more global perspective, consequently this theory can lose the perception of particular features within a phenomena and its concrete events. Theory of systems represents an intermediate level, but in its application one risks the loss of a more general perception, or a more specific view of particular aspects. Actors and scenarios’ foundations present a more specific approach, but for this reason, it does not offer a more global view as the chaos theory does.
In general the three main non-lineal models here presented do not exclude but include the quantitative approach for certain data. Moreover, the non-lineal methods create more richness in terms of generating hypotheses, not only proving them, as appears to be the case with the quantitative and rigorous methods for establishing the generality or restrictions of theoretical concepts.
Non-lineal methods are attempts to conform models with less distortion of reality since they can include but are not limited by the quantification of events, as the description of the relationships among main factors and their dynamic conditions. By using systems theory we can determine the degree of specificity in the relationships and the coherence or incoherence of the internal sub-systems involved. This objective can be achieved by implementing, under particular criteria, the concept of levels of integration and momentums. This "gradual" specification in the levels of concrete-abstract conception are less clear under the chaos and scenarios theory. However, these three main theories can establish specifically the "borders" and "time delimitations" of their phenomena to give its study the degree of concrete-abstract level that we want.
Finally, in terms of including quantitative data, the three methods here exposed have, to some extent, different approaches: a) Chaos theory can derive its main features from non lineal equations, such as the example of the origin of the butterfly effect; b) System concepts can include mathematical approaches from vector analysis as a set of variables to include, within sub-systems; relationships; levels of integration, and momentums; c) Actors and scenarios can include network analysis in order to determine the major trends in relationships, and the causal events for particular features and behavior of actors.
As mentioned in the introduction of this paper, the concrete case that I selected for the application of non-lineal methods is the economic adjustment plan implemented in Central America, which is the content of the second part.
Second Part: Application of non-lineal methodologies to the Economic Adjustment Measures Carried out in Central America
Under the current conditions Central American nations in general are facing three main concerns in terms of their sustainable economic development: a) To maintain economic growth with acceptable degree of economic redistribution of wealth, to avoid the increasing number of people living in poverty; b) To encourage the industrial production and to make significant and important changes in the structure of export products; and c) To promote economic integration in two senses, into NAFTA and other Latin American regional blocks like the Andean Group and CARICOM, and to reinforce in specific ways the Central American Common Market.
The economic development problem in Central America with those three objectives needs to deal with several actors, their relationships in social and political power and the external influence. The main characteristics from these elements are:
a) Actors. There are actors who have the real political power such as the business sectors in each country, the armies of the region, organized groups of the Central American financial system, and the political elites. Each one of them has its own version of the economic and social crises according to its particular interests. Other actors as secondary axis of power include the labor movement, the popular organization, the institutional bureaucracy, and churches;
b) Relationships. The main actors have among them relationships in terms of their economic and social interest. These relationships act on the international and national level, and they are going to shape the future characteristics of the economic policy in general, and the particular measures to carry out economic reactivation, integrational efforts and special international agreements.
c) External influence. This element is very important due to the marginal conditions of the Central American nations, their external vulnerability, and the identification of the elite especially with European and the economic centers in the United States. External influence is crucial for market and financial reasons in order to implement the mechanisms which in turn can guarantee economic growth and development.
According to the theory of chaos, the main points to underline are the trend in the regional blocks organizations, the incentives from the internal actors, and how these motivations can change over time. The problem that we can have is the too general perspective from the chaos theory, unless we are going to apply it to several subsystems.
Considering the sub-systems, we need to introduce in our analysis the system's theory, because we are going to pay more attention to the "internal" forces and their interaction with the external conditions and limitations. In doing that, we face the problem that the subsystems needed to be considered in two complementary ways: a) The actors inside as primary elements of the system behavior, and the relationships as the cohesive forces; and b) The more global conditions, as outcomes, that is the scenarios.
The scenarios theory can be useful to take into consideration those two aspects, and for this reason scenarios theory is in this case a good analytical tool. However, we need to pay attention to the role of the public policy makers and public policy analysts who can become real axis of power. One example of this situation can be seen in the conditionality terms to carry out the economic adjustment program due to the intervention of international organizations, mainly the International Monetary Found. In this case, public policy analysts of this institution can become actors with real power determining the principal aspects of the political agendas to be carry out by Central American countries.
With the considerations from the last paragraph I will summarize which I consider two additional variables to the policy making process.
Criteria for decision making process: Two additional variables
According to Stokey (Public Policy Analysis), there are four main "groups-perspectives" to classify actors who in turn are affecting the public policy making process: a) Rationalist, who have a global view, specific methods to determine the public policy to evaluate or implement, and who are mainly technicians with high level of qualification; b) Technicians, who are trained personal with specific and concrete tasks to carry out, and who are frequently criticized for their narrowness in the public policy making process. They are technicians as practitioners, who can have a deep level of specialization in their particular field; c) Incrementalists, who are often politicians; they are very concern for keeping their positions and for this reason they generally adopt conservative attitudes; d) Reformists, usually are referred to groups of pressure in the outside "field" of the traditional and official politics.
I think that these four categories are important in a public policy analysis. In order to have a better picture concerning the current and future conditions for particular scenarios, and for applying two additional variables of analysis -levels of power and spheres of action-, the case of the economic development tasks in the Central American countries will be useful.
a) Levels of political power. They are "quotas of power" that some actors or group of actors can have. There are levels of power from structural or permanent conditions -i.e. the armies in the region, with the exception of Costa Rica-, that have more political power, more than almost any other institution in each country. Other levels of power can derived from specific and non-permanent situations, such as the power of elected officials. They are going to be in power only for some years, or they can be deprived of their positions, for example when coups d'état occurs;
b) Spheres of action. I will call spheres of action the systemic (external) and sub-systemic (national or internal) circumstances. In fact I am supporting the idea of two main systems: the external and the internal, but with the condition that these two systems are in fact very interconnected in the economic development problems in Central America. It is difficult to consider an "objective" division between these two systems and their limitation is a conventional one;
I think it is possible to have more theoretical instruments within the non-lineal methods, to evaluate the current conditions in Central America. These theoretical elements will be useful to formulate and carry out the main economic tasks in Central America. These elements are related to: a) perspective-groups for public policy analysis (rationalists, technicians, Incrementalists, and Reformists); b) Levels of power, structural and non-permanent; and c) spheres of action, systemic and sub-systemic. Social and economic problems can be seen in a deeper manner using the aforementioned concepts, for example:
a) "Rationalist" and "technicians" of the International Monetary Found -IMF- acting with considerable structural level of power, from the systemic (external) conditions, can have real quota of power in determining national agendas. That means that they can establish the economic and fiscal measures to be carried out within the terms of conditionality. This situation took effect when Central American nations obtained financial resources from the IMF (Costa Rica 1979, 1986, 1993, 1995; Honduras 1982, 1993; El Salvador 1989, 1991; Nicaragua 1990, 1992, 1995; Guatemala 1984, 1986, 1989, 1993);
b) Armies in the region can support progressive taxes usually when the financial resources can be useful to provide basic services and basic economic infrastructure in the countryside. This situation has been especially evident in the cases of Nicaragua, Guatemala, and El Salvador. All these countries have had armed internal conflicts of considerable magnitude. In these cases actors with real structural power in the sub-systemic level are the crucial factors to carry out measures that business sectors are reluctant to accept;
c) Regional economic and commercial forces from NAFTA have direct influence in the future of the Central American Market. The marginal situation of the economies of the Central American nations, due to the structure of export products and size of their internal effective demand (27 million of people, with 75 percent of them living in poverty), do not allow these nations to reactivate their own economic integrational process. This condition is reinforced by the existence of NAFTA. In fact Costa Rica, and El Salvador are negotiating some kind of individual agreement with the United States to have preferential commercial treatment. In this case, business and political elites as actors from the systemic (external) level with structural and non-permanent levels of power, are the determinant forces of economic integrational measures in the Central American region;
These two variables that I underline here -levels of power and spheres of action- are complementary ones to the specific and concrete study of actors and their relationships in particular conditions. These two variables help us to see interactions that we cannot see using only actors, systems and sub-systems as unit of analysis.
Levels of power and spheres of action are additional elements that already can be studied in the methods derived from the chaos theory, the simulation with the scenario's perspective, and the theory of systems. They can be useful tools to formulate, evaluate, and monitor results, or for establishing some trends in public policy analysis.
In any case, there is no particular formula to claim as valid for every case in public policy analysis. The concrete method to use will depend on the conditions of analysis, available data, sources of information, specific outcomes, and above else, the fundamental characteristics of the phenomena under study.
Even when in some cases we can use mathematical and statistical methods, these quantitative procedures do not substitute the qualitative analysis and the interpretative stages of every study. Mathematics will provide tools, and elements for a better research, but it is always necessary to interpret results, and to establish the limitations and generalizations from our conclusions.
Pittsburgh, July 1999
Return to Sincronia, Summer 1999
Return to Sincronia General Index