Jay Forrester, circa 1975
Introduction
The nation exhibits a growing sense of futility as it repeatedly attacks deficiencies in our social system while the symptoms continue to worsen. Legislation is debated and passed with great promise and hope. But many programs prove to be ineffective. Results often seem unrelated to those expected when the programs were planned. At times programs cause exactly the reverse of desired results.
It is now possible to explain how such contrary results can happen. There are fundamental reasons why people misjudge the behavior of social systems. There are orderly processes at work in the creation of human judgment and intuition that frequently lead people to wrong decisions when faced with complex and highly interacting systems. Until we come to a much better understanding of social systems, we should expect that attempts to develop corrective programs will continue to disappoint us.
The purpose is to leave with its readers a sense of caution about continuing to depend on the same past approaches that have led to our present feeling of frustration and to suggest an approach which can eventually lead to a better understanding of our social systems and thereby to more effective policies for guiding the future.
A New Approach to Social Systems
It is my basic theme that the human mind is not adapted to interpreting how social systems behave. Our social systems belong to the class called multi-loop nonlinear feedback systems. In the long history of evolution it has not been necessary for man to understand these systems until very recent historical times. Evolutionary processes have not given us the mental skill needed to properly interpret the dynamic behavior of the systems of which we have now become a part.
In addition, the social sciences have fallen into some mistaken “scientific” practices which compound man’s natural shortcomings. Computers are often being used for what the computer does poorly and the human mind does well. At the same time the human mind is being used for what the human mind does poorly and the computer does well. Even worse, impossible tasks are attempted while achievable and important goals are ignored.
Until recently there has been no way to estimate the behavior of social systems except by contemplation, discussion, argument, and guesswork. To point a way out of our present dilemma about social systems, I will sketch an approach that combines the strength of the human mind and the strength of today’s computers. The approach is an outgrowth of developments over the last 40 years, in which much of the research has been at the Massachusetts Institute of Technology.
The concepts of feedback system behavior apply sweepingly from physical systems through social systems. The ideas were first developed and applied to engineering systems. They have now reached practical usefulness in major aspects of our social systems.
I am speaking of what has come to be called industrial dynamics. The name is a misnomer because the methods apply to complex systems regardless of the field in which they are located. A more appropriate name would be system dynamics. In our own work, applications have been made to corporate policy, to the dynamics of diabetes as a medical system, to the growth and stagnation of an urban area, and most recently to world dynamics representing the interactions of population, pollution, industrialization, natural resources, and food. System dynamics, as an extension of the earlier design of physical systems, has been under development at M.I.T. since 1956. The approach is easy to understand but difficult to practice. Few people have a high level of skill; but preliminary work is developing all over the world. Some European countries and especially Japan have begun centers of education and research.
Computer Models of Social Systems
People would never attempt to send a space ship to the moon without first testing the equipment by constructing prototype models and by computer simulation of the anticipated space trajectories. No company would put a new kind of household appliance or electronic computer into production without first making laboratory tests. Such models and laboratory tests do not guarantee against failure, but they do identify many weaknesses which can then be corrected before they cause full-scale disasters.
Our social systems are far more complex and harder to understand than our technological systems. Why, then, do we not use the same approach of making models of social systems and conducting laboratory experiments on those models before we try new laws and government programs in real life? The answer is often stated that our knowledge of social systems is insufficient for constructing useful models. But what justification can there be for the apparent assumption that we do not know enough to construct models but believe we do know enough to directly design new social systems by passing laws and starting new social programs? I am suggesting that we now do know enough to make useful models of social systems. Conversely, we do not know enough to design the most effective social systems directly without first going through a model-building experimental phase. But I am confident, and substantial supporting evidence is beginning to accumulate, that the proper use of models of social systems can lead to far better systems, laws, and programs.
It is now possible to construct in the laboratory realistic models of social systems. Such models are simplifications of the actual social system but can be far more comprehensive than the mental models that we otherwise use as the basis for debating governmental action.
Before going further, I should emphasize that there is nothing new in the use of models to represent social systems. Each of us uses models constantly. Every person in his private life and in his business life instinctively uses models for decision making. The mental image of the world around you which you carry in your head is a model. One does not have a city or a government or a country in his head. He has only selected concepts and relationships which he uses to represent the real system. A mental image is a model. All of our decisions are taken on the basis of models. All of our laws are passed on the basis of models. All executive actions are taken on the basis of models. The question is not to use or ignore models. The question is only a choice among alternative models.
The mental model is fuzzy. It is incomplete. It is imprecisely stated. Furthermore, within one individual, a mental model changes with time and even during the flow of a single conversation. The human mind assembles a few relationships to fit the context of a discussion. As the subject shifts so does the model. When only a single topic is being discussed, each participant in a conversation employs a different mental model to interpret the subject. Fundamental assumptions differ but are never brought into the open. Goals are different and are left unstated. It is little wonder that compromise takes so long. And it is not surprising that consensus leads to laws and programs that fail in their objectives or produce new difficulties greater than those that have been relieved.
For these reasons we stress the importance of being explicit about assumptions and interrelating them in a computer model. Any concept or assumption that can be clearly described in words can be incorporated in a computer model. When done, the ideas become clear. Assumptions are exposed so they may be discussed and debated.
But the most important difference between the properly conceived computer model and the mental model is in the ability to determine the dynamic consequences when the assumptions within the model interact with one another. The human mind is not adapted to sensing correctly the consequences of a mental model. The mental model may be correct in structure and assumptions but, even so, the human mind — either individually or as a group consensus — is most apt to draw the wrong conclusions. There is no doubt about the digital computer routinely and accurately tracing through the sequences of actions that result from following the statements of behavior for individual points in the model system. This inability of the human mind to use its own mental models is clearly shown when a computer model is constructed to reproduce the assumptions held by a single person. In other words, the model is refined until it is fully agreeable in all its assumptions to the perceptions and ideas of a particular person. Then, it usually happens that the system that has been described does not act the way the person anticipated. Usually there is an internal contradiction in mental models between the assumed structure and the assumed future consequences. Ordinarily the assumptions about structure and internal motivations are more nearly correct than are the assumptions about the implied behavior.
The kind of computer models that I am discussing are strikingly similar to mental models. They are derived from the same sources. They may be discussed in the same terms. But computer models differ from mental models in important ways. The computer models are stated explicitly. The “mathematical” notation that is used for describing the model is unambiguous. It is a language that is clearer, simpler, and more precise than such spoken languages as English or French. Its advantage is in the clarity of meaning and the simplicity of the language syntax. The language of a computer model can be understood by almost anyone, regardless of educational background. Furthermore, any concept and relationship that can be clearly stated in ordinary language can be translated into computer model language.
There are many approaches to computer models. Some are naive. Some are conceptually and structurally inconsistent with the nature of actual systems. Some are based on methodologies for obtaining input data that commit the models to omitting major concepts and relationships in the psychological and human reaction areas that we all know to be crucial. With so much activity in computer models and with the same terminology having different meanings in the different approaches, the situation must be confusing to the casual observer. The key to success is not in having a computer; the important thing is how the computer is used. With respect to models, the key is not to computerize a model, but instead to have a model structure and relationships which properly represent the system that is being considered.
I am speaking here of a kind of computer model that is very different from the models that are now most common in the social sciences. Such a computer model is not derived statistically from time-series data. Instead, the kind of computer model I am discussing is a statement of system structure. It contains the assumptions being made about the system. The model is only as good as the expertise which lies behind its formulation. Great and correct theories in physics or in economics are few and far between. A great computer model is distinguished from a poor one by the degree to which it captures more of the essence of the social system that it presumes to represent. Many mathematical models are limited because they are formulated by techniques and according to a conceptual structure that will not accept the multiple-feedback-loop and nonlinear nature of real systems. Other models are defective because of lack of knowledge or deficiencies of perception on the part of the persons who have formulated them.
But a recently developed kind of computer modeling is now beginning to show the characteristics of behavior of actual systems. These models explain why we are having the present difficulties with our actual social systems and furthermore explain why so many efforts to improve social systems have failed. In spite of their shortcomings, models can now be constructed that are far superior to the intuitive models in our heads on which we are now basing national social programs.
This approach to the dynamics of social systems differs in two important ways from common practice in social sciences and government. There seems to be a common attitude that the major difficulty is shortage of information and data. Once data is collected, people then feel confident in interpreting the implications. I differ on both of these attitudes. The problem is not shortage of data but rather our inability to perceive the consequences of the information we already possess. The system dynamics approach starts with the concepts and information on which people are already acting. Generally these are sufficient. The available perceptions are then assembled in a computer model which can show the consequences of the well-known and properly perceived parts of the system. Generally, the consequences are unexpected.
Counterintuitive Nature of Social Systems
Our first insights into complex social systems came from our corporate work. Time after time we have gone into a corporation which is having severe and well-known difficulties. The difficulties can be major and obvious such as a falling market share, low profitability, or instability of employment. Such difficulties are known throughout the company and by anyone outside who reads the management press. One can enter such a company and discuss with people in key decision points what they are doing to solve the problem. Generally speaking we find that people perceive correctly their immediate environment. They know what they are trying to accomplish. They know the crises which will force certain actions. They are sensitive to the power structure of the organization, to traditions, and to their own personal goals and welfare. In general, when circumstances are conducive to frank disclosure, people can state what they are doing and can give rational reasons for their actions. In a troubled company, people are usually trying in good conscience and to the best of their abilities to solve the major difficulties. Policies are being followed at the various points in the organization on the presumption that they will alleviate the difficulties. One can combine these policies into a computer model to show the consequences of how the policies interact with one another. In many instances it then emerges that the known policies describe a system which actually causes the troubles. In other words, the known and intended practices of the organization are fully sufficient to create the difficulty, regardless of what happens outside the company or in the marketplace. In fact, a downward spiral develops in which the presumed solution makes the difficulty worse and thereby causes redoubling of the presumed solution.
The same downward spiral frequently develops in government. Judgment and debate lead to a program that appears to be sound. Commitment increases to the apparent solution. If the presumed solution actually makes matters worse, the process by which this happens is not evident. So, when the troubles increase, the efforts are intensified that are actually worsening the problem.
Dynamics of Urban Systems
Our first major excursion outside of corporate policy began in February, 1968, when John F. Collins, former mayor of Boston, became Professor of Urban Affairs at M.I.T. He and I discussed my work in industrial dynamics and his experience with urban difficulties. A close collaboration led to applying to the dynamics of the city the same methods that had been created for understanding the social and policy structure of the corporation. A model structure was developed to represent the fundamental urban processes. The proposed structure shows how industry, housing, and people interact with each other as a city grows and decays. The results are described in my book Urban Dynamics, and some were summarized in Technology Review (April, 1969, pp. 21-31).
I had not previously been involved with urban behavior or urban policies. But the emerging story was strikingly similar to what we had seen in the corporation. Actions taken to alleviate the difficulties of a city can actually make matters worse. We examined four common programs for improving the depressed nature of the central city. One is the creation of jobs as by bussing the unemployed to the suburbs or through governmental jobs as employer of last resort. Second was a training program to increase the skills of the lowest-income group. Third was financial aid to the depressed city as by federal subsidy. Fourth was the construction of low-cost housing. All of these are shown to lie between neutral and detrimental almost irrespective of the criteria used for judgment. They range from ineffective to harmful judged either by their effect on the economic health of the city or by their long-range effect on the low-income population of the city.
The results both confirm and explain much of what has been happening over the last several decades in our cities. In fact, it emerges that the fundamental cause of depressed areas in the cities comes from excess housing in the low-income category rather than the commonly presumed housing shortage. The legal and tax structures have combined to give incentives for keeping old buildings in place. As industrial buildings age, the employment opportunities decline. As residential buildings age, they are used by lower-income groups who are forced to use them at a higher population density. Therefore, jobs decline and population rises while buildings age. Housing, at the higher population densities, accommodates more low-income urban population than can find jobs. A social trap is created where excess low-cost housing beckons low-income people inward because of the available housing. They continue coming to the city until their numbers so far exceed the available income opportunities that the standard of living declines far enough to stop further inflow. Income to the area is then too low to maintain all of the housing. Excess housing falls into disrepair and is abandoned. One can simultaneously have extreme crowding in those buildings that are occupied, while other buildings become excess and are abandoned because the economy of the area cannot support all of the residential structures. But the excess residential buildings threaten the area in two ways — they occupy the land so that it cannot be used for job-creating buildings, and they stand ready to accept a rise in population if the area should start to improve economically.
Any change which would otherwise raise the standard of living only takes off the economic pressure momentarily and causes the population to rise enough that the standard of living again falls to the barely tolerable level. A self-regulating system is thereby at work which drives the condition of the depressed area down far enough to stop the increase in people.
At any time, a near-equilibrium exists affecting population mobility between the different areas of the country. To the extent that there is disequilibrium, it means that some area is slightly more attractive than others and population begins to move in the direction of the more attractive area. This movement continues until the rising population drives the more attractive area down in attractiveness until the area is again in equilibrium with its surroundings. Other things being equal, an increase in population of a city crowds housing, overloads job opportunities, causes congestion, increases pollution, encourages crime, and reduces almost every component of the quality of life.
This powerful dynamic force to re-establish an equilibrium in total attractiveness means that any social program must take into account the eventual shifts that will occur in the many components of attractiveness. As used here, attractiveness is the composite effect of all factors that cause population movement toward or away from an area. Most areas in a country have nearly equal attractiveness most of the time, with only sufficient disequilibrium in attractiveness to account for the shifts in population. But areas can have the same composite attractiveness with different mixes in the components of attractiveness. In one area component A could be high and B low, while the reverse could be true in another area that nevertheless had the same total composite attractiveness. If a program makes some aspect of an area more attractive than its neighbor’s, and thereby makes total attractiveness higher momentarily, population of that area rises until other components of attractiveness are driven down far enough to again establish an equilibrium. This means that efforts to improve the condition of our cities will result primarily in increasing the population of the cities and causing the population of the country to concentrate in the cities. The overall condition of urban life, for any particular economic class of population, cannot be appreciably better or worse than that of the remainder of the country to and from which people may come. Programs aimed at improving the city can succeed only if they result in eventually raising the average quality of life for the country as a whole.
On Raising the Quality of Life
But there is substantial doubt that our urban programs have been contributing to the national quality of life. By concentrating total population, and especially low-income-population, in urban locations, undermining the strength and cohesiveness of the community, and making government and bureaucracy so big that the individual feels powerless to influence the system within which he is increasingly constrained, the quality of life is being reduced. In fact, if they have any effect, our efforts to improve our urban areas will in the long run tend to delay the concern about rising total population and thereby contribute directly to the eventual overcrowding of the country and the world.
Any proposed program must deal with both the quality of life and the factors affecting population. “Raising the quality of life” means releasing stress and pressures, reducing crowding, reducing pollution, alleviating hunger, and treating ill health. But these pressures are exactly the sources of concern and action aimed at controlling total population to keep it within the bounds of the fixed world within which we live. If the pressures are relaxed, so is the concern about how we impinge on the environment. Population will then rise further until the pressures reappear with an intensity that can no longer be relieved. To try to raise quality of life without intentionally creating compensating pressures to prevent a rise in population density will be self-defeating.
Consider the meaning of these interacting attractiveness components as they affect a depressed ghetto area of a city. First we must be clear on the way population density is, in fact, now being controlled. There is some set of forces determining that the density is not far higher or lower than it is. But there are many possible combinations of forces that an urban area can exert. The particular combination will determine the population mix of the area and the economic health of the city. I suggest that the depressed areas of most American cities are created by a combination of forces in which there is a job shortage and a housing excess. The availability of housing draws the lowest-income group until they so far exceed the opportunities of the area that the low standard of living, the frustration, and the crime rate counterbalance the housing availability. Until the pool of excess housing is reduced, little can be done to improve the economic condition of the city. A low-cost housing program alone moves exactly in the wrong direction. It draws more low-income people. It makes the area differentially more attractive to the poor who need jobs and less attractive to those who create jobs. In the new population equilibrium that develops, some characteristic of the social system must compensate for the additional attractiveness created by the low-cost housing. The counterbalance is a further decline of the economic condition for the area. But as the area becomes more destitute, pressures rise for more low-cost housing. The consequence is a downward spiral that draws in the low-income population, depresses their condition, prevents escape, and reduces hope. All of this is done with the best of intentions.
My paper, “Systems Analysis as a Tool for Urban Planning” from a symposium in October, 1969, at the National Academy of Engineering, suggests a reversal of present practice in order to simultaneously reduce the aging housing in our cities and allocate land to income-earning opportunities. The land shifted to industry permits the “balance of trade” of the area to be corrected by allowing labor to create and export a product to generate an income stream with which to buy the necessities of modern life from the outside. But the concurrent reduction of excess housing is absolutely essential. It supplies the land for new jobs. Equally important, the resulting housing shortage creates the population-stabilizing pressure that allows economic revival to proceed without being inundated by rising population. This can all be done without driving the present low-income residents out of the area. It can create upward economic mobility to convert the low-income population to a self-supporting basis.
The first reaction of many people to these ideas is to believe that they will never be accepted by elected officials or by residents of depressed urban areas. But some of our strongest support and encouragement is coming from those very groups who are closest to the problems, who see the symptoms first-hand, who have lived through the failures of the past, and who must live with the present conditions until enduring solutions are found.
Over the last several decades the country has slipped into a set of attitudes about our cities that are leading to actions that have become an integral part of the system that is generating greater troubles. If we were malicious and wanted to create urban slums, trap low-income people in ghetto areas, and increase the number of people on welfare, we could do little better than follow the present policies. The trend toward stressing income and sales taxes and away from the real estate tax encourages old buildings to remain in place and block self-renewal. The concessions in the income tax laws to encourage low-income housing will in the long run actually increase the total low-income population of the country. The highway expenditures and the government loans for suburban housing have made it easier for higher-income groups to abandon urban areas than to revive them. The pressures to expand the areas incorporated by urban government, in an effort to expand the revenue base, have been more than offset by lowered administrative efficiency, more citizen frustration, and the accelerated decline that is triggered in the annexed areas. The belief that more money will solve urban problems has taken attention away from correcting the underlying causes and has instead allowed the problems to grow to the limit of the available money, whatever that amount might be.
Characteristics of Social Systems
I turn now to some characteristics of social systems that mislead people. These have been identified in our work with corporate and urban systems and in more recent work that I will describe concerning the worldwide pressures that are now enveloping our planet.
First, social systems are inherently insensitive to most policy changes that people select in an effort to alter the behavior of the system. In fact, a social system tends to draw our attention to the very points at which an attempt to intervene will fail. Our experience, which has been developed from contact with simple systems, leads us to look close to the symptoms of trouble for a cause. When we look, we discover that the social system presents us with an apparent cause that is plausible according to what we have learned from simple systems. But this apparent cause is usually a coincident occurrence that, like the trouble symptom itself, is being produced by the feedback-loop dynamics of a larger system. For example, as already discussed, we see human suffering in the cities; we observe that it is accompanied (some think caused) by inadequate housing. We increase the housing and the population rises to compensate for the effort. More people are drawn into and trapped in the depressed social system. As another example, the symptoms of excess population are beginning to overshadow the country. These symptoms appear as urban crowding and social pressure.
Rather than face the population problem squarely we try to relieve the immediate pressure by planning industry in rural areas and by discussing new towns. If additional urban area is provided it will temporarily reduce the pressures and defer the need to face the underlying population question. The consequence, as it will be seen 25 years hence, will have been to contribute to increasing the population so much that even today’s quality of life will be impossible.
A second characteristic of social systems is that all of them seem to have a few sensitive influence points through which the behavior of the system can be changed. These influence points are not in the location where most people expect. Furthermore, if one identifies in a model of a social system a sensitive point where influence can be exerted, the chances are still that a person guided by intuition and judgment will alter the system in the wrong direction. For example in the urban system, housing is a sensitive control point but, if one wishes to revive the economy of a city and make it a better place for low-income as well as other people, it appears that the amount of low-income housing must be reduced rather than increased. Another example is the world-wide problem of rising population and the disparity between the standards of living in the developed and the underdeveloped countries, an issue arising in the world system to be discussed in the following paragraphs. But it is beginning to appear that a sensitive control point is the rate of generation of capital investment.
And how should one change the rate of capital accumulation? The common answer has been to increase industrialization, but recent examination suggests that hope lies only in reducing the rate of industrialization. This may actually help raise quality of life and contribute to stabilizing population.
As a third characteristic of social systems, there is usually a fundamental conflict between the short-term and long-term consequences of a policy change. A policy which produces improvement in the short run, within five to ten years, is usually one which degrades the system in the long run, beyond ten years. Likewise, those policies and programs which produce long-run improvement may initially depress the behavior of the system. This is especially treacherous. The short run is more visible and more compelling. It speaks loudly for immediate attention. But a series of actions all aimed at short-run improvement can eventually burden a system with long-run depressants so severe that even heroic short-run measures no longer suffice. Many of the problems which we face today are the eventual result of short-run measures taken as long as two or three decades ago.
The great challenge for the next several decades will be to advance understanding of social systems in the same way that the past century has advanced understanding of the physical world. (1981) Jay Forrester
I address several social concerns: population trends; quality of urban life; policies for urban growth; and the unexpected, ineffective, or detrimental results often generated by government programs. The field of system dynamics now can explain how such contrary results happen.
Fundamental reasons cause people to misjudge behavior of social systems. Orderly processes in creating human judgment and intuition lead people to wrong decisions when faced with complex and highly interacting systems. Until we reach a much better public understanding of social systems, attempts to develop corrective programs for social troubles will continue to be disappointing. We caution against continuing to depend on the same past approaches that have led to present feelings of frustration. New methods developed over the last 30 years will lead to a better understanding of social systems and thereby to more effective policies for guiding the future.
A new approach to social system behavior
The human mind is not adapted to interpreting how social systems behave. Social systems belong to the class called multi-loop nonlinear feedback systems. In the long history of evolution it has not been necessary until very recent historical times for people to understand complex feedback systems.
Evolutionary processes have not given us the mental ability to interpret properly the dynamic behavior of those complex systems in which we are now imbedded. The social sciences, which should be dealing with the great challenges of society, have instead retreated into small corners of research. Various mistaken practices compound our natural mental shortcomings. Computers are often being used for what computers do poorly and the human mind does well. At the same time the human mind is used for what the human mind does poorly and computers do well. Furthermore, impossible tasks are attempted while achievable and important goals are ignored.
Until recently, no way to estimate the behavior of social systems existed except by contemplation, discussion, argument, and guesswork. As a way out of the present dilemma, I will sketch here an approach that combines the strength of the human mind and the strength of today’s computers.
Concepts of feedback system behavior apply sweepingly from physical systems through social systems. Feedback system ideas were first developed and applied to engineering systems. Understanding closed-loop (feedback) systems has now reached practical usefulness in social systems.
COMPUTER MODELS OF SOCIAL SYSTEMS
People would never send a spaceship to the moon without first testing prototype models and making computer simulations of anticipated trajectories. No company would put a new household appliance or airplane into production without first making laboratory tests. Such models and laboratory tests do not guarantee against failure, but they do identify many weaknesses which can be corrected before they cause full-scale disasters.
Social systems are far more complex and harder to understand than technological systems. Why then do we not use the same approach of making models of social systems and conducting laboratory experiments before adopting new laws and government programs?
The customary answer assumes that our knowledge of social systems is not sufficient for constructing useful models. But what justification can there be for assuming that we do not know enough to construct models of social systems but believe we do know enough to directly redesign social systems by passing laws and starting new programs? I suggest that we now do know enough to make useful models of social systems. Conversely, we do not know enough to design the most effective social policies directly without first going through a model-building experimental phase.
Substantial supporting evidence is accumulating that proper use of models of social systems can lead to far better systems, laws, and programs. Realistic laboratory models of social systems can now be constructed. Such models are simplifications of actual systems, but computer models can be far more comprehensive than the mental models that would otherwise be used. Before going further, please realize that there is nothing new in the use of models to represent social systems. Each of us uses models constantly. Every person in private life and in business instinctively uses models for decision making. The mental images in one’s head about one’s surroundings are models.
One’s head does not contain real families, businesses, cities, governments, or countries. One uses selected concepts and relationships to represent real systems. A mental image is a model. All decisions are taken on the basis of models. All laws are passed on the basis of models. All executive actions are taken on the basis of models. The question is not to use or ignore models. The question is only a choice among alternative models. Mental models are fuzzy, incomplete, and imprecisely stated. Furthermore, within a single individual, mental models change with time, even during the flow of a single conversation. The human mind assembles a few relationships to fit the context of a discussion. As debate shifts, so do the mental models. Even when only a single topic is being discussed, each participant in a conversation employs a different mental model to interpret the subject.
Fundamental assumptions differ but are never brought into the open. Goals are different but left unstated. It is little wonder that compromise takes so long. And even when consensus is reached, the underlying assumptions may be fallacies that lead to laws and programs that fail. The human mind is not adapted to understanding correctly the consequences implied by a mental model. A mental model may be correct in structure and assumptions but, even so, the human mind–either individually or as a group consensus–is apt to draw the wrong implications for the future.
Inability of the human mind to use its own mental models becomes clear when a computer model is constructed to reproduce the assumptions contained in a person’s mental model. The computer model is refined until it fully agrees with the perceptions of a particular person or group. Then, usually, the system that has been described does not act the way the people anticipated. There are internal contradictions in mental models between assumed structure and assumed future consequences. Ordinarily assumptions about structure and internal governing policies are more nearly correct than are the assumptions about implied behavior.
By contrast to mental models, system dynamics simulation models are explicit about assumptions and how they interrelate. Any concept that can be clearly described in words can be incorporated in a computer model. Constructing a computer model forces clarification of ideas. Unclear and hidden assumptions are exposed so they may be examined and debated. The primary advantage of a computer simulation model over a mental model lies in the way a computer model can reliably determine the future dynamic consequences of how the assumptions within the model interact with one another.
There need be no doubt about a digital computer accurately simulating the actions that result from statements about the structure and policies in a model. In some ways, computer models are strikingly similar to mental models. Computer models are derived from the same sources; they may be discussed in the same terms. But computer models differ from mental models in important ways. Computer models are stated explicitly. The “mathematical” notation used for describing the computer models is unambiguous.
Computer simulation language is clearer, simpler, and more precise than spoken languages. Computer instructions have clarity of meaning and simplicity of language syntax. Language of a computer model can be understood by almost anyone, regardless of educational background. Furthermore, any concept that can be clearly stated in ordinary language can be translated into computer-model language.
There are many approaches to computer models. Some are naive. Some are conceptually inconsistent with the nature of actual systems. Some are based on methodologies for obtaining input data that commit the models to omitting major relationships in the psychological and human areas that we all know to be crucial. With so much activity in computer models and with the same terminology having different meanings in the different approaches, the situation is confusing to a casual observer. The key to success is not in having a computer; the important thing is how the computer is used. With respect to models, the key is not to computerize a model, but, instead, to have a model structure and decision-making policies that properly represent the system under consideration.
I am speaking here of system dynamics models—the kind of computer models that are only now becoming widely used in the social sciences. System dynamics models are not derived statistically from time-series data. Instead, they are statements about system structure and the policies that guide decisions. Models contain the assumptions being made about a system. A model is only as good as the expertise which lies behind its formulation. A good computer model is distinguished from a poor one by the degree to which it captures the essence of a system that it represents. Many other kinds of mathematical models are limited because they will not accept the multiple-feedback-loop and nonlinear nature of real systems.
On the other hand, system dynamics computer models can reflect the behavior of actual systems. System dynamics models show how difficulties with actual social systems arise, and demonstrate why so many efforts to improve social systems have failed. Models can be constructed that are far superior to the intuitive models in people’s heads on which national social programs are now based.
System dynamics differs in two important ways from common practice in the social sciences and government. Other approaches assume that the major difficulty in understanding systems lies in shortage of information and data. Once data is collected, people have felt confident in interpreting the implications. I differ on both of these attitudes. The problem is not shortage of data but rather inability to perceive the consequences of information we already possess. The system dynamics approach starts with concepts and information on which people are already acting.
Generally, available information about system structure and decision-making policies is sufficient. Available information is assembled into a computer model that can show behavioral consequences of well-known parts of a system. Generally, behavior is different from what people have assumed.
COUNTERINTUITIVE NATURE OF SOCIAL SYSTEMS
Our first insights into complex social systems came from corporate work. Time after time we went into corporations that were having severe and well known difficulties. The difficulties would be obvious, such as falling market share, low profitability, or instability of employment. Such difficulties were known throughout the company and were discussed in the business press.
One can enter a troubled company and discuss what people see as the causes and solutions to their problems. One finds that people perceive reasonably correctly their immediate environments. They know what they are trying to accomplish. They know the crises which will force certain actions. They are sensitive to the power structure of the organization, to traditions, and to their own personal goals and welfare. When interviewing circumstances are conducive to frank disclosure, people state what they are doing and can give rational reasons for their actions.
In a troubled company, people are usually trying in good conscience and to the best of their abilities to help solve the major difficulties. Policies are being followed that they believe will alleviate the difficulties. One can combine the stated policies into a computer model to show the consequences of how the policies interact with one another. In many instances it emerges that the known policies describe a system which actually causes the observed troubles. In other words, the known and intended practices of the organization are sufficient to create the difficulties being experienced.
Usually, problems are blamed on outside forces, but a dynamic analysis often shows how internal policies are causing the troubles. In fact, a downward spiral can develop in which the presumed solutions make the difficulties worse and thereby cause greater incentives to redouble the very actions that are the causes of trouble. The same downward spiral frequently develops in government. Judgment and debate lead to a program that appears to be sound. Commitment increases to the apparent solution. If the presumed solution actually makes matters worse, the process by which degradation happens is not evident. So, when the troubles increase, the efforts are intensified that are actually worsening the situation.
CHARACTERISTICS OF SOCIAL SYSTEMS
Many characteristics of social systems mislead people. Behavior that people do not anticipate appears in corporate and urban systems and in world-wide pressures now enveloping the planet. Three counterintuitive behaviors of social systems are especially dangerous.
First, social systems are inherently insensitive to most policy changes that people choose in an effort to alter the behavior of systems. In fact, social systems draw attention to the very points at which an attempt to intervene will fail. Human intuition develops from exposure to simple systems. In simple systems, the cause of a trouble is close in both time and space to symptoms of the trouble. If one touches a hot stove, the burn occurs here and now; the cause is obvious.
However, in complex dynamic systems, causes are often far removed in both time and space from the symptoms. True causes may lie far back in time and arise from an entirely different part of the system from when and where the symptoms occur. However, the complex system can mislead in devious ways by presenting an apparent cause that meets the expectations derived from simple systems. A person will observe what appear to be causes that lie close to the symptoms in both time and space—shortly before in time and close to the symptoms.
However, the apparent causes are usually coincident occurrences that, like the trouble symptom itself, are being produced by the feedback-loop dynamics of a larger system. For example, human suffering in cities is accompanied (some think caused) by inadequate housing. As a result, housing is increased and population rises to defeat the effort. More people are trapped in the depressed urban system.
Second, social systems seem to have a few sensitive influence points through which behavior can be changed. These high-influence points are not where most people expect. Furthermore, when a high-influence policy is identified, the chances are great that a person guided by intuition and judgment will alter the system in the wrong direction. For example, in an urban system, housing is a sensitive control point but, if one wishes to make the city a better place for low-income as well as other people, it appears that low-income housing should be reduced rather than increased. Another example lies in the world-wide problem of rising population and the disparity between the standards of living in the developed and the underdeveloped countries.
System dynamics models suggest sensitive control points for increasing the world-wide quality of life exist in the rate of generation of capital investment and in food production, but that expansion of industrialization and food output are the counterproductive directions, both should be restrained. The common answer to world distress has been to increase industrialization and food production, but hope for long-term improvements probably lies in reducing emphasis on both. Contrary to intuitive expectations, the opposite of present practice may actually raise the quality of life and contribute to stabilizing population.
Third, social systems exhibit a conflict between short-term and long-term consequences of a policy change. A policy that produces improvement in the short run is usually one that degrades a system in the long run. Likewise, policies that produce long-run improvement may initially depress behavior of a system.
This is especially treacherous. The short run is more visible and more compelling. Short-run pressures speak loudly for immediate attention. However, sequences of actions all aimed at short-run improvement can eventually burden a system with long-run depressants so severe that even heroic short-run measures no longer suffice. Many problems being faced today are the cumulative result of short-run measures taken in prior decades.
A GLOBAL PERSPECTIVE
It is certain that resource shortage, pollution, crowding, disease, food failure, war, or some other equally powerful force will limit population and industrialization if persuasion and psychological factors do not. Exponential growth cannot continue forever. At present population growth rates, there would remain only one square yard per person in less than 400 years. Our greatest challenge is to guide the transition from growth to equilibrium. There are many possible mechanisms for limiting growth. That current growth rates of population and industrialization will stop is inevitable. Unless we choose favorable processes to limit growth, the social and environmental systems by their internal processes will choose for us. The natural mechanisms for terminating exponential growth appear the least desirable. Unless the world understands and begins to act soon, civilization will be overwhelmed by forces we have created but can no longer control.
ATTRACTIVE POLICIES CAN CREATE DISASTERS
We may not be fortunate enough to run gradually out of natural resources. Science and technology may very well find ways to use the more plentiful metals and atomic energy so that resource depletion does not intervene. If so, the way then remains open for another growth-resisting pressure to arise In other words, the standard of living is sustained with a lower drain on the expendable and irreplaceable resources. But the picture is even less attractive! By not running out of resources, population and capital investment are allowed to rise until a pollution crisis is created. Pollution then acts directly to reduce birth rate, increase death rate, and depress food production. Population, which according to this simple model peaks at the year 2030, has fallen to one-sixth of the peak population within an interval of 20 years—a world-wide catastrophe of a magnitude never before experienced. Should it occur, one can speculate on which sectors of the world population will suffer most. It is quite possible that the more industrialized countries (which are the ones that would have caused such a disaster) would be the least able to survive such a disruption to environment and food supply. Industrialized countries might take the brunt of the collapse.
The way in which a solution to one trouble creates a new problem has defeated many past governmental programs and will continue to do so until more effort is devoted to understanding the dynamic behavior of social systems.
A NEW FRONTIER
It is now possible to take hypotheses about separate parts of a social system, combine them in a computer model, and learn the consequences. The hypotheses may at first be no more correct than the ones we are using in our intuitive thinking. But the process of computer modeling and model testing requires these hypotheses to be stated explicitly. The model comes out of the hazy realm of mental models into unambiguous model statements to which all have access. Assumptions can then be checked against all available information and can be rapidly improved.
The great uncertainty with mental models arises from inability to anticipate the consequences of interactions between parts of a system. This uncertainty about future dynamic implications of assumptions in a model is totally eliminated in computer models. Given a stated set of assumptions, the computer traces the resulting consequences without doubt or error. Computer simulation is a powerful procedure for clarifying issues. It is not easy. Results will not be immediate.
We are on the threshold of a great new era in human pioneering. What we do today affects our future many decades hence. If we follow intuition and the fallacies embedded in mental models, the trends of the past will continue into deepening difficulty. If we set appropriate research and education programs, which are now possible, we can expect a far sounder basis for future action.
THE NATION’S ALTERNATIVES
Our greatest challenge now is handling the transition from growth to equilibrium. For a thousand years, tradition has encouraged and rewarded growth. Folklore and success stories praise growth and expansion. But growth is not the path for an unlimited future. Many present stresses in society arise from pressures that accompany the transition from growth into equilibrium. However, the pressures thus far in cities are minor compared to those which are approaching. Population pressures and economic forces in a city that was reaching equilibrium have in the past been able to escape to new land areas.
Escape is becoming less possible. Until now we have had, in effect, an inexhaustible supply of farm land and food-growing potential, but now we are reaching a critical point where, all at the same time, population is overrunning productive land, agricultural land is almost fully employed for the first time, the rise in population is putting more demand on food supplies, and urbanization is pushing agriculture out of the fertile areas into marginal lands.
For the first time demand is rising into a condition where supply will begin to fall while need increases. The crossover from plenty to shortage can occur abruptly. The fiscal and monetary policies of a country form a complex dynamic system of the kind I have been discussing. It is clear that the United States has no established policies to guide interactions between government, growth, unemployment, and inflation. The need to develop long-term policies becomes ever more urgent as the country moves for the first time from a history of growth into the turbulent pressures accompanying the transition from growth to one of the many possible kinds of equilibrium. We need to choose and work toward a desirable kind of equilibrium before we arrive at a point where the system imposes its own choice of regrettable consequences.
In a hierarchy of systems, a conflict exists between goals of a subsystem and welfare of the broader system. The conflict is seen in an urban system. The goal of a city is to expand and to try to raise its quality of life. But growth policies increase population, industrialization, pollution, and demands on food supply. The broader social systems of a country and the world require that goals of urban areas be curtailed and that pressures from such curtailment become high enough to keep urban areas and population within bounds that are satisfactory to the larger system of which cities are a part. If this nation continues to pursue traditional urban goals, the result will deepen distress of the country as a whole and eventually deepen the crisis in cities themselves. We may be at a point where higher pressures in the present are necessary if insurmountable pressures are to be avoided in the future.
Evolutionary processes have not given us the mental ability to interpret properly the dynamic behavior of those complex systems in which we are now imbedded. The social sciences, which should be dealing with the great challenges of society, have instead retreated into small corners of research. Various mistaken practices compound our natural mental shortcomings. Society becomes frustrated as repeated attacks on deficiencies in social systems lead only to worse symptoms. Legislation is debated and passed with great hope, but many programs prove to be ineffective. Results are often far short of expectations. Because dynamic behavior of social systems is not understood, government programs often cause exactly the reverse of desired results. Jay Forrester (1971)
I have given a glimpse of the nature of multi-loop feedback systems, a class to which social systems belong. I have shown how these complex systems mislead people because intuition has been formed by experienced from simple systems from which we expect behavior very different from that actually possessed by complex systems. The United States is still pursuing programs that will be even more frustrating and futile than many of the past.
Visits: 178