Back to main document about Beyond the Limits.
"Global models" are made to investigate something happening over the entire world. Many global models exist. Global models have differing purposes, and their creators have different perspectives and use different methods. This reader is about one global model, "World3," which is used in the book Beyond the Limits. This model is one of several for which information is being made available an interdisciplinary project. The project is creating software that teaches about Earth systems from the perspectives of social, physical, and biological systems. Also available is The Gaia Hypothesis and Daisyworld.
The World3 global model is a tool to think about and plan for the future. Policy makers, researchers, and nearly everyone else have ideas about the world's future -- about what the future will be like, and why. You undoubtedly have your own ideas. Your ideas are a sort of global model in your head, a "mental model." The World3 model of Beyond the Limits was constructed from the mental models of researchers. But World3 is more than a mental model. Its assumptions are explicit, so everyone can read about, analyze, discuss, and criticize those assumptions. Because World3 exists as equations on a computer, the computer can think through the model's consequences completely. It does not forget to account for parts of the model, as people often forget. Use the model to help examine what you are quietly thinking, assuming, and hoping. Use it to help you probe and revise your own mental model of the world's future and of how to improve the future.
World3 was designed to understand how human society will approach the world's changing carrying capacity, and what conditions or policies will increase the chance of a smooth approach to the capacity. When first introduced in 1972, Beyond the Limits' predecessor The Limits to Growth caused a storm of controversy about its grim portrayal of the future: dramatic collapse resulting from uncontrolled growth. Yet the point of both studies is not to predict doom, but to plan a better future.
Use the program to try worldwide strategic plans. Plan a strategy, and make any necessary changes to the model. For example, you might want to try developing new farming technologies. But don't start the model yet! First think through what you expect to happen. Draw graphs or make notes about how population, industrial output, and other variables will change over time. Why will they change this way? Then start the model. Do the results differ from your expectations? If so, why? Adjust your mental model appropriately. If you disagree with World3's assumptions, or if you want to try a different strategy, change the model and repeat the process.
It's best to change variables one at a time. This can give you the most thorough understanding of why the model behaves as it does. Start by (1) using this program's default setting (scenario 2 in Beyond the Limits), (2) setting the world as you think it is most likely to be, or (3) setting the world as you wish it would be. Then make changes and run the model in ways that you think will help you learn.
As you come to understand both your own model and the Beyond the Limits model, you will have taken a step toward understanding and changing your future and the future of Earth.
Back to contents.
The book Beyond the Limits discusses real-world data about world population and industrial growth, and about several limits that growth may encounter. This discussion is critical to understanding how the world might change in future. If you cannot read Beyond the Limits, then read on for a condensed discussion.
Back to contents.
Conclusion: If we plan to grow the world's population much larger, we'd better ensure -- somehow -- (1) that we can use a lot more land for farming than we now use, and (2) that land yields increase. Unless we ensure these changes, or unless we can come up with a workable, inexpensive wonder-technology fix, there will be a world food shortage IF the world's population grows much larger. Just when that shortage would begin is hard to say, and depends on what we achieve.
Back to contents.
Conclusion: Many of the world's countries will confront limits on water supplies within several decades. Some countries will be able to cut back on water use with relative ease. Others will need more innovative solutions. Less growth of population and of industry would mean less demand for water, hence fewer water crises. Desalination and detoxification of water might be used by wealthy regions.
Back to contents.
Conclusion: If the world wants to keep its forests, which are valuable resources for ecological sustainability, biodiversity, building materials, fuel, and other reasons, then the world needs to cut back drastically on deforestation.
Back to contents.
Conclusion: Whether and when fossil fuel shortages become a problem depends on how efficiently we manage to extract resources and what resource pools we discover underground. Shifts of resource-extraction from some nations to others could contribute to political turbulence. Energy efficiency is the most cost-effective way to avoid shortages, and it is desireable in any case because it can save money without detrimental changes. Technological improvements to reduce extraction costs are and have been ongoing. Renewable energy sources such as solar and wind power are promising, and they deserve more R&D effort to increase their cost-effectiveness.
Back to contents.
Conclusion: There may be problems obtaining various materials from the ground, as exponentially growing extraction causes the grade of ore to fall below the cutoff grade. Recycling of materials might be able to solve this problem. This would require the development of recycling techniques and programs not presently used.
Back to contents.
Conclusion: Pollution is an important problem; how important is unknown. It is conceivable that pollution could cause large-scale breakdown of ecosystems including the natural systems that destroy or remove certain pollutants; at present such a breakdown seems unlikely but we don't know the details to say for sure. Pollution increases with population growth and with industrialization. Because of everyday problems, ignoring drastic dangers, it seems to make sense in industrialized nations to decrease many forms of pollution, as is being done by filtering or stopping emissions from factories and automobiles. Underdeveloped nations usually cannot afford the price of such pollution controls. Developed nations need to do as much as they can to help underdeveloped nations to adopt pollution control technologies at minimal cost. This will benefit not only the underdeveloped nations, but also the developed nations, because pollution has a global impact.
Back to contents.
Conclusion: World population growth is slowing, but estimates of the stabilized population level range around 7-10 billion, and the estimates could be wrong. Population growth drives requirements on food and water. It also drives industrial growth, which drives requirements on all the topics dicussed earlier. Population growth can be slowed by a change in attitudes. One of the best ways to change attitudes is to empower women through education and jobs. Family planning programs in the developing world need the support of the industrialized world so that they can be successful. These programs will benefit both the developing world and the industrialized world by reducing negative impacts of population. In theory, population growth can help a developing country by creating more producers and enlarging the economy, but in practice this generally does not bear out -- instead, food, housing, and consumer goods per-capita decrease.
Back to contents.
Conclusion: Exponentially-increasing industry means exponentially-increasing demand for energy, materials, water, and wood, and exponentially increasing pollution. To avoid detrimental impacts from these demands and from pollution, one or both of two changes need to happen. (1) Industrial growth needs to slow or stop. (2) We have to come up with technological and innovative solutions to decrease the energy, materials, water, and wood needed by industry, and to decrease the pollution generated by industry. If we are going to use technological and innovative solutions rather than slowing or stopping industrial growth, we'd better act fast. Because in a matter of decades -- perhaps a half-century or a century, perhaps sooner -- the detrimental impacts will become urgent problems.
Back to contents.
This section lists some key ideas from Beyond the Limits. Use that book to become familiar with these ideas. Ask yourself in what ways the ideas are important.
Back to contents.
Many global simulation models have been constructed. They have addressed different purposes, been constructed in different ways, and espoused different viewpoints. For a comparative look at global models, their purposes, and their construction, see these two books:
1. Meadows, Richardson, & Bruckman. Groping in the Dark: The First Decade of Global Modeling. Wiley: 1982. This book compares many global modeling studies, and it discusses the point of global modeling and methodological issues. It comes from the sixth of a series of IIASA (International Institute for Applied Systems Analysis) conferences on global modeling (though it is not a typical book of conference papers) and it maps out the ways in which scientists who do global modeling agree or disagree.
2. Barney, Kreutzer, & Garrett. Managing a Nation: The Microcomputer Software Catalog. Westview Press: 1991. This book describes computer software to help in planning the affairs of a nation. In addition to global models, the book discusses single-issue models (e.g. models of agriculture, demography, or energy).
Back to contents.
The World2 model, predecessor to World3, appeared in Jay Forrester's 1971 book World Dynamics. The World3 model was the basis for the 1972 book The Limits to Growth. These models generated considerable interest and controversy. Beyond the Limits uses the World3 model to update and reassess the findings of The Limits to Growth, twenty years after the original study. Since Beyond the Limits is recent and has received less attention, it has received less criticism than the other books. Therefore, this section largely discusses previous criticisms that are still relevant to Beyond the Limits.
The authors of World2 and World3 have responded to most criticisms, but they have communicated their responses to few people. This section attempts to clear up the old lack of communication. Where a response to a criticism exists, that response is discussed along with the criticism. Some new criticisms come from this section's author, Simons.
Back to contents.
Some critics claim that the debate about World3 is optimists arguing against pessimists. World3's authors are supposedly pessimists who see bad possible futures: e.g., starvation, resource exhaustion, and collapse of the global economy. The optimists supposedly see good possible futures: e.g., high-tech farming, solar power and recycling, and high-efficiency production. Critics often state that reality is somewhere in between the two extremes. In reality, the critics say, there will be some of what the optimists say, and some of what the pessimists say. The future will not be as bad as the World3 model supposedly predicts, but neither will it match people's wishful dreams.
Actually, World3's authors try to account for the most critical forces that help to determine future patterns of the world's population and industrial growth. In addition to accounting for "bad" forces, they account for "good" forces of technological development and changing human behavior. They do not claim to know how technologies will develop, nor how human behavior will change. Instead, they ask, "If we humans can develop such-and-such sorts of technologies, and if we change our behaviors in such-and-such ways, what sorts of futures might happen?" They use the World3 model to probe many possible patterns of technological development and of behavioral change.
Some critics have argued that World3's representation of technology is inappropriate, or that the model ought to account for good forces that it does not yet include. These criticisms will be discussed later. If indeed the World3 model leaves out good factors that it should have accounted for, or if it treats such factors inappropriately, then World3 may be overly pessimistic. But in general, the optimist-pessimist view is incorrect, because World3's authors intended to account for all the key forces relevant to changes in population and industrial growth. World3's authors do not merely restate Malthus' gloom-and-doom predictions about mass starvation. They also examine intentional and unintentional changes that could avert Malthus' predictions. The authors of Beyond the Limits intend that their conclusions come from a balanced viewpoint, not from pure pessimism.
Back to contents.
Many critics of World Dynamics and The Limits to Growth stated that the World3 model did not, and ought to, account for technology. In fact, both studies did account for technology. But the method by which World Dynamics represented technology may be inappropriate, and technology may have received short shrift in The Limits to Growth.
The World2 model of World Dynamics aggregated technology with industrial capital. That is, when the world's industries were large, they were assumed to have a lot of technology, and when small, to have little technology. In essence, the model viewed technology in terms of numbers of people with relevant science or engineering know-how, and in terms of the availability of the machines and tools used by scientists and engineers. World3 has another dimension of technology. It assumes that knowledge accumulates. The knowledge cannot go away if industry gets smaller. Knowledge accumulates about how to control pollution, decrease resource use, and increase crop yields. Once this knowledge is applied, according to World3, the world never reverts to older technology. (Presumably the world also accumulates knowledge in other areas of science and engineering, but the model only tracks technologies needed to think about population and industrial growth.)
In Science (11 August 1972), Robert Boyd described technologically optimistic changes to the World2 model. According to these assumptions, the world will no longer produce any pollution nor use any resources by about the year 2020. He showed that with the optimistic assumptions, World2 yields a future of unrestricted growth of population and industry. He concluded that how good or bad the model's results are depend upon whether its builders' assumptions are optimistic or pessimistic. He labeled Forrester's viewpoint as "Malthusian," and said that given his viewpoint it was no surprise that Forrester's model led to catastrophic results similar to those predicted by Malthus. The outcomes of simulation models, according to Boyd's article, are no better than the viewpoints of their authors.
Boyd did not test assumptions between what he called optimistic and pessimistic. Carrying out such tests shows that Forrester's conclusions are robust. If the assumptions are set as with Boyd's assumptions, but with a more realistic 50%, 75%, or even 90% reduction in pollution and natural resource use per manufactured good by about 2020, the assumptions still lead to global population and industrial collapse. As Forrester originally emphasized, curtailing the growth of population and industry can avert the crisis.
Other critics have suggested that the growth of technology could outpace the growth of problems caused by a large population and industry. If technology grows at an exponential rate that exceeds the growth rate for population and industry, technology could always be one step ahead of the problems. For example, if industry is expanding at 3% per year, then the resources consumed by industry, and the pollution generated by industry, might be decreased by technology at 4% per year. So far this has not been true, though perhaps the pattern will change and the world will develop such technologies at a pace far exceeding today's pace. The consequences of such high-paced technological improvement can be examined by altering the technology sections of the "Change the Model" section of this program. However, the model will assume that the technologies impose a high cost on industry; to change these costs using the program, see the Industrial Output section of the model diagrams.
Back to contents.
Economist Julian Simon has argued that population growth is good for society. The more people there are, he says, the better industry works, and the faster new technology is developed. Many demographers have argued against Simon. At least in developing countries, they say, more population seems to put greater demands on an already-limited food supply and an already-strained economy. Nevertheless, the mechanism discussed by Simon is likely correct even if his conclusion is wrong. With more minds to work on science and technology, it seems reasonable to expect that science and technology will improve faster -- though how much faster is debatable. The World3 modelers did not examine this effect, which could cause work on technology to increase over time. But you can examine the effect using the program. To do so, turn on the program's "game" mode. (Go to the "Run the Model" section and press Command-G. See this progam's manual for more information.) Start the simulation with the two technology controls set to the right of the middle, and move them farther right as the population increases, or back to the left if the population decreases.
World3 does take into account that people are needed to keep industry going. This is the reason for the "Jobs" section of the model.
Back to contents.
Many economists have argued that the World3 model fails because it does not model economic supply and demand. The argument seems to be that if an economic model does not analyze supply and demand, then it cannot be any good. In fact, World3 assumes that economic markets work perfectly. In the model, supply and demand always match. In the real world, of course, supply and demand don't always match. Thus World3 is slightly over-optimistic about economic markets. In any case, it is not practical to include in the model the supply and demand for most individual products. Supply and demand equilibrate with each other in a few weeks, months, or years, whereas the World3 model looks at what happens over a century. The necessary additions to the model, plus the need to analyze it in terms of weeks instead of years, would overcomplicate the model. Furthermore it is irrelevant to the model's purpose to know (and it is impossible to predict) what will be the prices of bananas, shoes, cars, toasters, etc.
In a few markets -- technology and nonrenewable resources -- the model does make supply and demand assumptions. Two assumptions about technology are that (1) developing a new technology takes time, and (2) people work hardest to develop a new technology when that technology is badly needed. (For details, see the relevant parts of the program's "Change the Model" section.) Two assumptions for nonrenewable resources are that (1) there is a limit to how much of one or more critical resources can be cost-effectively extracted from the Earth (though this limit can be nullified by changing the model to assume the limit is very high), and (2) when resources become hard to get the world's industry still gets them, but only by diverting resources from other forms of industry to mining, drilling, processing, and refining. These assumptions seem to match with empirical research on economic and technological change.
World3 does have flaws in how it describes different parts of the world economy and how those parts interrelate. One flaw in World3 is how it determines the cost of agricultural technology. In World3, the cost of a technology is given as a percentage reduction in industrial output. Regardless of whether half of industry or a thousandth of industry is needed to support agriculture, the percent reduction in industrial output is the same. (In the model, the fraction of industry used to support agriculture might vary between about a tenth and a quarter of industry.) A better approach would be to formulate the model so that a given amount of industrial output, not a given fraction of industrial output, goes to support agriculture and agricultural technologies. If world industrial output dropped to near or below the amount needed for agriculture, food production would get cut back, resulting in food shortages. One researcher has made this change in a revision to World3 which he calls "World4." The change does sometimes make a considerable difference in the model's behavior, though it does not alter the main conclusions discussed in Beyond the Limits. It is not yet possible to implement the change using this program. The problem with agricultural technology does not exist for pollution control technology, nor for resource efficiency technology. These technologies are used in all parts of industry, so it is appropriate to reduce the output of all parts of industry according to the costs of the technologies.
A second flaw in World3 is how it determines investment in new industrial capital. Investment in capital is what keeps industry running. This investment is in the form of new factories, new machinery, new training for workers, and new and repaired transportation and communication systems. To determine how much investment occurs at any point in time, World3 does the following. First, it calculates how much "stuff" industry is producing. The amount produced depends upon how much industrial capital exists, and it is decreased if there is a shortage of workers or if technologies such as pollution controls make production less efficient. Second, some of the "stuff" produced by industry goes to agriculture, some to human services, some to consumer goods, and some to digging up resources (metals, minerals, and fossil fuels) from the ground. Whatever "stuff" happens to be left over is used for new investment. Is this characterization appropriate? Does new investment really get determined this way? Probably not. Probably people will spend effort to invest in manufacturing industry if investment is needed, even if there are pressures to invest in the service industries or in agriculture. The model should be fixed to make industrial capital investment decisions more realistic. You can make some changes to this end via the Industrial Output section of the program's Model Diagrams--see the variables named "industrial equilibrium time" and "fraction of industrial output allocated to consumption variable." If you do not make these changes, you might expect that the model will sometimes show an industrial collapse when it should not do so, or at least before it should do so. The model can depict an industrial collapse caused by the high costs of several technologies. This collapse could lead to under-investment in agriculture, and hence mass starvation. Thus, the starvation in that particular example could be a consequence of the model's possibly-flawed treatment of investment decisions. In reality, industry might continue to invest in production and in agriculture, perhaps at the cost of fewer consumer goods and human services.
Back to contents.
To discuss some other common criticisms made by economists, this section dicusses a critique published in a standard intermediate economics text: Walter Nicholson's book Microeconomic Theory: Basic Principles and Extensions, 4th ed. (Chicago: Dryden, 1989), pp. 22-24.
Nicholson writes that industrial output in the model is proportional to the amount of resources available and to the quantity of capital that exists. This is incorrect. The effect of available resources on output is less than proportional until few resources remain, then more than proportional. Why? Industrial output (this by definition does not include extracted and refined resources, nor food) decreases only slightly as the world's supply of resources is depleted. The decrease in output occurs because the model assumes that extraction becomes more expensive as resources become more scarce, making the price of resources rise. At first the price of resources rises very little (technological improvements may even drive costs down for a time), but if resources are used until almost none remain, the cost of resource extraction rises much more steeply and the available resource quantity has a more than proportional effect on output.
Note that the term "price of resources" is used above as if all different kinds of resources (metals, minerals, and fuels) all had the same price. Of course they do not have the same price, but for purposes of simplicity, the model aggregates all resources into one bunch. More specifically, what matters to the model is the most critical set of resources. For what groups of resources that people dearly need is the world most likely to use up the resources' cheaply available supply? Is this most likely to happen with the metals used to make girders and vehicles and tools, with metals used to conduct electricity, with phosphorous (needed in fertilizer), with energy sources, or with other substances? Whichever group is most likely to run low (and is important for industry to keep running) is the critical resource that is aggregated together and called "nonrenewable resources." Resources that will certainly not run low are irrelevant to the model, which needs to know only whether some set of important resources may be depleted enough that a large part of worldwide industry must be devoted to extracting and processing that set of resources.
Nicholson states, "For economists, probably the major weakness of the Limits models was their failure to model prices." As an example he focuses on energy sources, though his arguments may be generalized to all kinds of nonrenewable resources. Criticisms related to not modeling prices were discussed in the previous section on Economic Supply and Demand, but here I address some specific details. Taking account of resource prices is important because (1) consumers may switch from a resource that is growing scarce (e.g. oil) to another that is more plentiful and hence cheaper (e.g. natural gas), and because (2) consumers may buy products (and industry make products) that are more resource-efficient than older products. Concern number 1 is taken care of in the model by aggregating together all substitutable critical resources, as discussed above. Thus the supply of nonrenewable energy sources could not run short until supplies run short for oil, natural gas, and coal. (Of course these energy sources might not be considered critical resources anyway if it is thought that they could be mostly replaced by renewable energy sources.) Concern number 2 involves two ways in which products may become more resource efficient. First it may be possible to use existing technology to improve on products, and second new technology may be developed. The World3 model groups together these two kinds of technical advance, which leads to some inaccuracy in analyzing the pace at which products become more resource-efficient, but this pace is not accurately known anyway. The model assumes no advances in resource efficiency unless that portion of the model is turned on. (The authors of Beyond the Limits do turn on such advances, comparing this to what would happen if no advances occurred. You also need to turn on these advances, using the Changes section of this program.) Also, the model ignores differences in the costs of applying these technologies, and instead assumes for simplicity that the cost to buy and make products that use fewer resources is the same regardless of whether those products depend upon technology that already exists or technology that must be newly developed. Thus while the World3 model is not fundamentally flawed in this respect, it is very much simplified.
Much of this discussion could be more specific--and perhaps more satisfactory--if the nonrenewable resource sector of the model were broken into parts that represent different types of resources. The reason why this was not done in the World3 model is that such changes would make the model very complicated to build and use (it is complicated enough already!) and because so little is known about the world's supply of different resources that the changes could make little or no improvement in the model's ability to predict the likelihood of different global futures. Instead, the model simply says that the world starts out with "a lot" of critical resources or "a few," as determined by the person using the model (see the Changes section of this program). With a lot of critical resources no limits on resources come into play, with few critical resources a limit does come into play (resource extraction costs become very high unless technology improves resource efficiency), and with a moderate amount of critical resources there may be a less drastic effect on industry because of slightly rising resource costs.
A revision of the World3 model being co-authored by Thomas Fiddaman does attempt somewhat to disaggregate the Nonrenewable Resources sector of the model into separate kinds of resources. And other models have examined various kinds of natural resources; see the book Managing a Nation mentioned in the chapter of this reader entitled Global Simulation Models.
Another concern of Nicholson's is that World3 "does not model labor input or possible improvements in the quality of this input (through education, for example). Instead, the authors focused only on workers as consumers of economic output." Indeed, the model assumes that the size of the world's population does not directly affect the pace of industrial growth; see the Impacts of Population Growth section of this chapter on Criticisms. However, if more population led to faster industrial growth, it would be more likely that the model would overshoot its "limits" to growth, by putting more demands on the world's resources and releasing more pollutants. Whether he is right in saying that improvements to the quality of labor are considered in the model is a more complicated issue. The people who built World3 did consider the effect of continual improvements in the quality (efficiency) of labor and of capital machinery, but made a reasoned decision that such improvements were roughly balanced by the fact that the ratio of labor to capital gradually falls (since wages rise, and since in the industrialized parts of the world industry grows faster than population). This is reflected in the variable "industrial capital-output ratio," which they assumed remains constant over time. (This variable can, however, change to reflect the costs of technologies developed to improve resource efficiency, control pollution, or increase crop yields.) Thus while this assumption of the model may be debated, it does not mindlessly ignore improvements in the inputs (capital and labor) to industry.
Finally, Nicholson complains that "the Limits authors adopted what another author termed a 'fruit fly' theory; that is, they assumed that increasing affluence in the world would bring increasing population growth." This is misleading. The model does assume that the short-term impact of affluence is to increase birth rates, because, according to the model, families change their perceived ability to support children but not their desired family sizes. Within about a decade, however, desired family sizes decrease, so that increasing affluence leads to decreasing population growth, as seems to be the case from available data.
Back to contents.
Some people have stated that The Limits to Growth was wrong because its predictions have not come true. This conclusion is inappropriate. The Limits to Growth predicted a potential collapse of global population and industry around mid-next century: perhaps around 2030, but later if technologies forestall the world's problems. It predicted no collapse if people acted soon to curtail growth of population and industry. Now (1996) is still too early to know whether the predictions of The Limits to Growth have come true.
Back to contents.
While The Limits to Growth and Beyond the Limits may have the right idea, their World3 model cannot analyze interaction among the world's different regions. They barely mention issues like: Might population and industrial collapse happen in some regions of the world before others? How is wealth distributed between nations, what would be an equitable distribution of wealth, and how can the world achieve an equitable distribution? What politics are involved? What political, economic, and social strategies should each nation use?
Some studies have attempted to analyze differences between regions. Mankind at the Turning Point, by Mihajlo Mesarovic and Eduard Pestel, uses a model similar to World3, but which divides the world into regions. The book summarizes its authors' conclusions from the project. Discussions of the model's inner workings took place at an IIASA (International Institute for Applied Systems Analysis) conference.
The "Globus" model was developed in the 1980s at a German research center. It analyzes international politics and economics over a forty-year period (1970 to 2010). A book explains how the model works and describes some of the ways it was used: The Globus Model: Computer Simulation of Worldwide Political and Economic Developments, edited by Stuart A. Bremer.
At least one person has stated that the Globus model proves that The Limits to Growth is wrong, because there is no population or industrial collapse in the Globus model that divides the world into regions. The conclusion is inappropriate. Globus does not look far enough into the future to compare it with World3, and it does not include the mechanisms that might cause a collapse of population and industry. It uses a pre-determined population forecast from the United Nations.
Back to contents.
Why does Beyond the Limits use a computer simulation model? Some critics argue that the model is unnecessary -- the book Beyond the Limits could have been written without mention of any computer model. The key ideas of the book would still come across to readers. The computer model is a way to gain legitimacy, these critics charge, by preying upon readers who are impressed by what they do not understand. They point out, correctly, that people tend to use computer models improperly, as electronic oracles that predict the future.
It is true that the book could have been written without mentioning a computer model. However, its authors might not have been able to reach their conclusions without creating a computer model. Computer modeling is a learning process. The model builders describe their assumptions about the world, in an explicit form which makes it easy to compare each other's ideas and empirical facts. The computer model uses assumptions that the model builders have agreed upon, that appear to match with reality. The computer can find the consequences of those assumptions, under a wide range of circumstances. If the model's results disagree with what the model builders expected, they can trace the disagreement either to an error in the model, or to an error in their expectations. Thus, depending on what people expect to happen in the world, they can learn through a process of revising the computer model, their assumptions, and their expectations. The outcomes of simulation models can be very different from their builders' initial viewpoints. With time, the builders' viewpoints and the models' results should come to agreement.
World3 should not be used to predict. It should be used to learn. It shows futures that could happen, and it helps users think about how to improve the future. It can also be used to explore people's assumptions, by comparing their assumptions to those in the model and to empirical fact, and by revising the model to match their assumptions and seeing what difference the changes cause -- and why -- in the model's results. Empirical facts relevant to the model are gathered together in the 1974 book Dynamics of Growth in a Finite World. This program's authors have done what they can to encourage proper use of the model, and to warn people of inappropriate uses. Please help out, by discouraging people from using the model as a "black box" to predict the future.
Back to contents.
A 1973 book, Models of Doom, remains the most comprehensive critique of the World3 model. The book is valuable for reflecting on the model's methods and assumptions. Many of the book's points have already been discussed, but this section addresses a few remaining points.
The book points out various simplifications, errors in formulation, and parts left out of the model that maybe ought to be included. Similar concerns are pointed out in this program. It is no surprise that the model is far from perfectly accurate. It is difficult to capture the truth when the scientific fields upon which the model builds are far from knowing the truth. What matters is, would fixing problems in the model change the conclusions one draws from it about the two questions it was designed to answer? This author has found no reason why such fixes would reverse the conclusions about these two questions.
Part of the Models of Doom critique was to say that human decisions and feedback processes ought to be taken into account in the World3 model, and that many such decisions and feedbacks are not considered (e.g., people might limit industrial growth if industrial pollution becomes intolerable, or decisions related to childbearing might change in ways not considered in the model). But in fact the World3 modelers have considered such decisions and feedbacks, by trying many different simulations with different parameter settings to reflect different possible human decisions. The feedbacks are or are not taken into account, depending on how various "policy parameters" are set. This reflects a goal of trying to understand how human behavior can make the world better, instead of trying to predict how humans will behave. If "critics" can come up with positive ways for people to change that are not considered in World3, that is wonderful. Hopefully you and others will be creative and resourceful at coming up with ways to make the world's population, industry, etc. sustainable, while maintaining a good quality of life for earth's citizens.
Models of Doom suggests "backcasting" as a way to check validity of the model. In fact, for reasons that involve the techniques of model computation, backcasting is not a valid way to check model validity, and well-established models in physics, economics, and other fields would often fail the same inappropriate test.
Reference: Cole, H.S.D., Christopher Freeman, Marie Jahoda, and K.L.R. Pavitt. Models of Doom: A Critique of The Limits to Growth. Universe Books, 1973.
Back to contents.
This program is a tool. As with any tool, it can be used beneficially or harmfully. Use it to help you think about the world's future. It can give you ideas. It can make you remember factors you might otherwise forget. It can help you revise your conceptions. Do not use it to tell fortunes. Use the World3 simulation model many times -- not just once -- to explore a range of possible futures and possible human actions. And, I recommend, keep the focus on the real world, which is after all the point of the model.
Included with the program is an educational workbook. For classrooms with one computer equipped with a projector, the workbook can guide classroom discussion. If the discussion takes a different path, that's fine. What the class does can be compared with the workbook, to make sure the class learns the major concepts in the workbook. The workbook can also be used as take-home assignments, or as a study guide for individuals learning on their own.
If you develop ideas for activities and questions, or if you have comments about the workbook or program, we'd like to hear your ideas. Send email to K.Simons@rhbnc.ac.uk. Or write to the following address: Re: Beyond the Limits software -- comments; P. Poole & K. Simons; Box 2310; Banff, Alberta T0L-0C0; CANADA.
Back to contents.
Meadows, Donella H., Dennis L. Meadows, and Jørgen Randers. Beyond the Limits. Chelsea Green Publishing (Post Mills, Vermont): 1992. You need to have read this book before using this Beyond the Limits computer program.
Meadows, Dennis L., et al. Dynamics of Growth in a Finite World. Wright- Allen Press: 1974. Now distributed by Productivity Press (Cambridge, Mass.). This book documents the World3 model. It is essential if you plan to explore the model in detail.
See the book Beyond the Limits, pages 267-274, for annotated bibliographies on the following topics:
Back to contents.
Copyright ©1997, 1996, 1995, 1994, 1993, 1992 by Kenneth L. Simons.
You may copy and distribute the Beyond the Limits program at no charge, provided that it is unmodified, and that you also distribute the program's accompanying materials.
This software is part of a project to create educational software about models of global society and environment, spearheaded by Peter Poole and Kenneth Simons. Also available from this project is the program The Gaia Hypothesis and Daisyworld. It is free.
To be on a mailing list for updates to software created by this project, contact Ken Simons.
The Beyond the Limits software was constructed by Kenneth Simons, Jennifer Newbury, and Aaron Young. A preliminary prototype for this program was created at MIT by Kenneth Simons, Michelle Bell, Peter Poole, Kevin Rathbun, and Erik Trimble.
Kenneth Simons thanks David Kreutzer, John Sterman, and the MIT System Dynamics Group for their support and encouragement; Tom Fiddeman and Dennis Meadows for their feedback; and Ann Bostrum and Baruch Fischhoff for their ideas. For work by Kenneth Simons: This material is based upon work supported under a National Science Foundation Graduate Fellowship. Any opinions, findings, conclusions or recommendations expressed in this publication are those of the author and do not necessarily reflect the views of the National Science Foundation.
Aaron Young thanks Dr. Joseph Devine and the Student-Defined Major program of the College of Humanities and Social Sciences of Carnegie Mellon University, for providing the opportunity to be involved in this project.
Back to contents.
Copyright ©1992-1997 by
Kenneth L. Simons.
Revised 8 September 1997