Sunday, 22 January 2017

The Collapse of Complex Societies, Part 2

In my last post I started a review of Joseph Tainter's book The Collapse of Complex Societies.

In that post, we saw Tainter develop his theory that complex societies collapse because of diminishing marginal returns on investments in complexity. He makes these observations:

  1. human societies are problem-solving organizations;
  2. sociopolitical system require energy or their maintenance;
  3. increased complexity carries with it increased costs per capita; and
  4. investments in sociopolitical complexity as a problem-solving response often reaches a point of declining marginal returns

I quoted Tainter on how, as marginal returns decline, societies find themselves spending so much on maintaining complexity that they don't have the resources to adequately responding to stress—various sorts of emergencies and disasters. I should have mentioned that this can also be looked at in terms of optimization and resilience. In order to minimize those decreasing marginal returns, complex systems are usually optimized as much as possible. That is to say, they are made as efficient as possible in terms of their intended output under ideal conditions. Unfortunately, this inevitable reduces their adaptability and resiliency. Such systems are very brittle and when even fairly minor things go wrong they tend to break (collapse) where a more resilient system would have been able to adapt and carry on.

He also makes it clear (to me, anyway) that adding complexity is our preferred strategy. Even when we reach the point where we are working as hard as we can to just maintain the status quo, we are very reluctant to simplify things. This would seem to imply that our path of progress from the caves to the stars is fated to be interrupted on occasion by collapse, and perhaps stopped altogether.

Of course, these ideas are very unwelcome to those of us who have been raised in the religion of progress. We have been taught that progress makes things work better, makes life more convenient and comfortable and holds great hope for the future. But very little is said about what it all costs. It is pretty clear to me that progress amounts to increasing complexity, even if some of that complexity is hidden. My dad's first car, for instance, (a Model T Ford) had a manual choke and manual spark advance and starting it took quite a bit of skill. Today the latest models have push button start which takes care of all those details and works pretty much every time. Much simpler you might say, but have a look under the hood and you'll find that it takes a lot of hidden complexity to achieve this "simplicity".

When I first read this book, I was unwilling to even accept the idea of decreasing marginal productivity. Even though Tainter offers extensive examples and statistics in Chapter 4 to show that this is indeed so. Whether it is agriculture, resource production, manufacturing, information processing, sociopolitical control, specialization or overall economic productivity—the law of decreasing returns does indeed apply. Indeed, it is one of the few areas of economics in which we have enough confidence to call a "law".

But what I couldn't understand is how you could have decreasing productivity when yields have been going up significantly. Of course, what we are talking about here is decreases in marginal productivity.

To use terminology from agriculture, the yields per acre have indeed improved. But the amount of additional (marginal) effort required has grown for each successive improvement. Effort means energy. Traditionally that meant muscle power, human or animal. Today we have achieved great reductions in the amount of muscle power required in agriculture, but only because we have learned to use energy (in the form of fossil fuels) directly to drive machines or manufacture fertilizers and pesticides. With all this, agriculture has become an increasingly complex endeavour which consumes more energy than ever before. And over the last couple of decades yields have started to level off despite heroic measures to improve them.

Another illustration of this is the development of the coal-based economy in England, covered by Tainter in chapter 4 of this book. One commonly hears the switchover to coal extolled as a great leap forward, due to the greater BTUs per unit weight to be had from burning coal. Unfortunately, that viewpoint suffers from a certain amount of tunnel vision.

Jumps in population at around AD 1300, 1600 and in the eighteenth century, led to intensification in agriculture and industry and as land was increasingly deforested to provide fuel and agricultural space, basic heating, cooking and manufacturing needs could no longer be met by burning wood. A shift to reliance on coal began, gradually and with apparent reluctance. Coal was definitely a fuel source of secondary desirability, being more expensive to obtain and distribute than wood as well as being more dirty and polluting... Mining of coal from the ground was more costly than obtaining a quantity of wood of equal heating value and became ever more costly as the most accessible reserves were depleted. Mines had to be sunk ever deeper, until groundwater problems became a serious problem. Ultimately the steam engine was developed to pump water out of mines, using some of the coal as a power source...

The increased costliness per unit of thermal value in the initial shift from wood to coal is apparent, but unfortunately good data on returns to energy investment are not available before the recent period. Modern data not only illustrate the trend quantitatively, but indicates that the process of declining marginal returns is continuing.

That was written in 1988. It is interesting to note, from our vantage point in 2017, that coal mining in the UK peaked in 1913 and finally came to an end in 2015.

Another aspect of increasing complexity is that, in order to minimize those decreasing marginal returns, complex systems are usually optimized as much as possible. That is to say, they are made as efficient as possible in terms of their intended output. Unfortunately, this inevitable reduces thier adaptability and resiliency. Such systems are very brittle and when even fairly minor things go wrong they tend to break (collapse) where a more resilient system would have been able to adapt and carry on.

In Chapter 5, Tainter takes a close look at three examples of collapse to see if his law of decreasing marginal productivity works as an explanation. These are The Western Roman Empire, The Classic Maya of the Southern Lowlands and The Chacoan Society of the American Southwest. Tainter goes into quite a bit of detail about each of these societies, but I'll just share his conclusions:

  1. In each of the cases examined, the costliness of complexity increased over time while the benefits to the population declined.
  2. In each substantial increased costs occurred late, shortly before the collapse, and these were imposed on a population already weakened by the previous pattern of decreasing marginal returns
  3. For Rome and Maya, population leveled off or declined before the collapse and the well-being of most people deteriorated. This seems to have come about from the demands of supporting such complex systems. It is not currently known whether something similar happened in the Chacoan case, but it is noteworthy that the number of Outliers participating in this system dropped prior to the final collapse. Quite possibly, Outlier communities, whose participation could not be enforced (unlike the Roman and Mayan cases), withdrew from the network before declining marginal productivity adversely affected their local populations.
  4. For the Maya and Chacoans, subsequent abandonment of their territories, and the lack of a substantial reoccupation by agricultural peoples, suggests that there was environmental deterioration during the period of growth. This may indicate that pressures of population on resources had more to do with the Mayan and Chacoan collapses than with that of Rome. The Roman case is very different, for the later Empire was decidedly under populated.
  5. In each case, people on the periphery (the northern European barbarians, the northern Maya, and the Western and Eastern Pueblos) rose to prominence after the older society had collapsed.

None of these cases can be completely understood by the explanations commonly advanced for them.

  1. The fall of Rome was not due to barbarians, for the Empire was economically, organizationally, and militarily stronger than the besiegers. And it was not due to internal weakness, for the Empire remained essentially intact for a period of several hundred years. Rome's collapse was due to the excessive costs imposed on the agricultural population to maintain a far-flung empire in a hostile environment.
  2. The fall of the Maya was not due to a peasant revolt, for peasants supported this civilization for over 1000 years. It was not due to invasions, for which there is unclear evidence and uncertain causality, nor to agricultural deterioration, for the evidence of agricultural intensification indicates that the Maya were fully capable of increasing the productivity of their environment. The collapse of the Maya civilization was due to the burdens of an increasingly costly society borne by an increasingly weakened population. Peasant dissatisfaction, foreign pressures, internal conflict, or an agricultural crisis may have provided a final, insurmountable challenge, but such a challenge was effective only because the Maya were following a course that made them vulnerable to collapse.
  3. The Chacoan collapse was not due to drought or environmental deterioration, for these were factors which the Chacoans were technically capable of dealing with, and indeed had previously done so. The regional populations of the San Juan Basin chose not to continue participating in the Chacoan network, nor to rise to the challenge of the final drought because the costs of doing so had grown too high in comparison to the advantages conferred. Collapse and migration were economically preferable.

This chapter began with the observation that the framework for explaining collapse could probably not be subjected to a formal, quantitative test. The alternative was to investigate three cases in detail, asking whether the framework developed in Chapter 4 helps us to understand why these societies collapsed. The results lead us to answer the question affirmatively: the collapse of the Western Roman Empire, the Southern Lowland Maya and the Chacoan society can be understood as responses to declining marginal returns on investment in complexity.

In Chapter 6 Tainter summarizes what he has been saying so far, and then considers the idea that collapse may not universally be a catastrophe:

...under a situation of declining marginal returns collapse may be the most appropriate response.

What may be a catastrophe to administrators (and later observers) need not be to the bulk of the population... It may only be among those members of a society who have neither the opportunity nor the ability to produce primary food resources that the collapse of administrative hierarchies is a clear disaster. Among those less specialized, severing the ties that link local groups to a regional entity is often attractive. Collapse then is not intrinsically a catastrophe. It is a rational, economizing process that may well benefit much of the population.

Tainter concludes by considering the implications for contemporary societies. And here I believe he finds himself in something of a state of denial. Or it may be that he felt a wholehearted endorsement of the idea that collapse lies in our future would have just been too much for his audiences to swallow. Tainter was well aware of the response experienced by The Limits to Growth.

He acknowledges that his findings certainly point to the possibility of collapse of our complex modern society and that such a collapse could be devastating because much of the population does not have the opportunity or ability to produce primary food resources. In addition to mentioning existential threats like comet strikes, he lists a number of scenarios for contemporary collapse:

  • nuclear war and associated climatic changes
  • increasing atmospheric pollution, leading to ozone depletion, climatic changes, saturation of global circulation patterns, and similar disasters
  • depletion of critical industrial resources
  • general economic breakdown, brought on by such things as unrepayable national and international debts, disruptions to fossil fuel availability, hyperinflation, and the like.

He even acknowledges that industrialism will someday have to deal with resource depletion and its own wastes. The major question is how far off the day is. And he admits that patterns of declining marginal returns can be observed in at least some contemporary industrial societies in the following areas: agriculture; mineral and energy production; research and development; investment in health and education; government, military and industrial management; productivity of GNP for producing new growth; and some elements of improved technical design.

He identifies two opposing reactions to all this: economists who believe the challenges we faces are all solvable economic dilemmas (to be solved by economic growth), and environmentalists who believe that stimulating economic growth can only hasten the inevitable crash and hold that we should be aiming for economic undevelopment (what is known today as "degrowth"). He says that both sides are ignoring key historical factors.

He proceeds to do a good job of debunking techno-optimism, talking clearly about what is wrong with the principle of infinite substitutability:

One problem with the principle of infinite substitutability is that it does not apply, in any simple fashion, to investments in organizational complexity. Sociopolitical organization, as we know, is a major arena of declining marginal returns, and one for which no substitute product can be developed. Economies of scale and advances in information-processing technology do help lower organizational costs, but ultimately these too are subject to diminishing returns.

A second problem is that the principle of infinite substitutability is, despite its title, difficult to apply indefinitely. A number of perceptive scientists, philosophers, and economists have shown that the marginal cost of research and development... have grown so high it is questionable whether technological innovation will be able to contribute as much to the solution of future problems as it has to past ones...

It is not that R&D cannot potentially solve the problems of industrialism. The difficulty is that to do so will require an increasing share of GNP. The principle of infinite substitutability depends on energy and technology. With diminishing returns to investment in scientific research, how can economic growth be sustained? The answer is that to sustain growth, resources will have to be allocated from other sectors of the economy into science and engineering. The result will likely be at least a temporary decline in the standard of living... The allocation of greater resources to science of course is nothing new, merely the continuation of a two century-old trend. Such investment, unfortunately, can never yield a permanent solution, merely a respite from diminishing returns.

But then, in my opinion, he goes astray. He says that historical collapses were all of complex societies functional in isolation. Today's world is full of complex societies in competition with each other, which (he asserts) changes things.

Any nation vulnerable to collapse will have to pursue one of three options: (1) absorption by a neighbour or some larger state; (2) economic support by a dominant power, or by an international funding agency; or (3) payment by the support population of whatever costs are needed to continue complexity, however detrimental the marginal return. A nation today can no longer unilaterally collapse, for if any national government disintegrates its population and territory will be absorbed by some other.

His comments on degrowth are limited to this:

Here is the reason why proposals for economic undevelopment, for living in balance on a small planet, will not work. Given the close link between economic and military power, unilateral economic deceleration would be equivalent to and as foolhardy as, unilateral disarmament. We simply do not have the option to return to a lower economic level, at least not a rational option. Competition among political peers drives increased complexity and resource consumption regardless of costs, human or ecological.

I do not wish to suggest by this discussion that any major power would be quickly in danger of collapse were it not for this situation. Both the primary and secondary world powers have sufficient economic strength to finance diminishing returns well into the future.

It is amusing (and sad) that this was written very shortly before the collapse of the Soviet Union, during which the rest of the world stood by and declined to finance the USSR any further. Evidently, there was insufficient economic strength to finance diminishing returns.

Since then we have witnessed numerous cases where the world powers have stood by and allowed an assortment of small countries to undergo something very similar to collapse. Why? Because it didn't make economic sense to do anything else. Off the top of my head I would point to Cuba, Yugoslavia, Haiti, Somalia, Rwanda, Greece and Venezuela.

The major powers are simply no longer ready to finance diminishing returns in those cases where there is no profit in it for them. The phenomena of "throwing them to the wolves", both in the case of small economically insignificant countries, and inconvenient groups and individuals within countries, is becoming standard procedure. Help from international funding agencies has turned out to be the kiss of death for developing countries. And people are very unwilling to support governments that call for austerity as a path to economic recovery (rightly so).

Tainter goes on to say (as have so many others) that a new energy subsidy beyond fossil fuels will be necessary to finance declining marginal returns, and even that will only be a temporary fix. He believed that the lack of a power vacuum and the resulting competitive spiral have given us a temporary reprieve in which to search for that energy subsidy.

There is both cause for optimism and pessimism in the current situation. We are in a curious position where competitive interactions force a level of investment, and a declining marginal return that might ultimately lead to collapse except that the competitor who collapses first will simply be dominated or absorbed or dominated by the survivor... fact industrial societies are subject to the same principles that caused earlier societies to collapse. If civilization collapses again, it will be from failure to take advantage of the current reprieve, a reprieve paradoxically both detrimental and essential to our anticipated future.

Sorry, Mr. Tainter—in the event, that's not the way it turned out. Whatever reprieve there was in the closing days of the twentieth century has slipped between our fingers. In the twenty-first century, those who collapse are largely abandoned by those striving to avoid a similar fate.

This just goes to show you how difficult making predictions is, especially about the future. But none of that takes away in the least from the main message of this book, which I would say is rock solid—human societies use complexity to solve problems and because of declining marginal returns, those solutions are always temporary and often lead to collapse. This principle should be kept very much in mind by those of us who are trying understand the events unfolding around us in the world today.

I recommend reading the book, but there are also a quite a few videos featuring Joseph Tainter on YouTube.

Sunday, 15 January 2017

The Collapse of Complex Societies, Part 1

In a few weeks I'll be starting a series of posts called Political Realities: Collapse Step by Step, looking into how the collapse of industrial society is likely to proceed. The title is a reference to a series of posts I did a while back, entitled A Political Fantasy which looked at what could be done to improve our situation if the political will existed. In the upcoming series I'll be looking at what is going to happen instead. Before I get there, though, I'll be doing reviews of a few books that will help us understand some features of this collapse.

In this post and the next one, we'll be looking at Dr. Joseph Tainter's Collapse of Complex Societies. It was published in 1988, and is a classic among us kollapsniks.

In his first chapter, Tainter tells us that collapse is a political process and that a society has collapsed when it displays a rapid, significant loss of an established level of sociopolitical complexity. He goes on to describe the following characteristics of a collapsed society:

  • a lower degree of stratification and social differentiation
  • less economic and occupational specialization of individuals, groups and territories
  • less centralized control; that is, less regulation and integration of diverse economic and political groups by elites
  • less behavioral control and regulation
  • less investment in the epiphenomena of complexity, those elements that define the concept of "civilization": monumental architecture, artistic and literary achievements and the like
  • less flow of information between individuals, between political and economic groups and between a center and its periphery
  • less sharing, trading and redistribution of resources
  • less overall coordination and organization of individual and groups
  • a smaller territory integrated within a single political unit

I have a few points to add to that list:

  • collapse is always political because the entity which is collapsing (a complex society) is essentially political
  • "rapid" is a relative term—if you are living through events that will later be seen as happening rapidly, it will probably seem to you like it is taking forever.
  • as a collapse is happening, very few people see it as such. What at the time are optimistically seen as temporary setbacks are later clearly seen as stages of collapse.
  • when a society is in collapse, efforts to prevent changes to less complex forms may do more harm than good and gracefully accepting the inevitable may be beneficial. Later on in the book Tainter shows us why this is so.

He goes on to look at collapses in history. I think he may have been trying to establish how common a process it is, for the benefit of those less familiar with the subject. He makes a few remarks about the collapse of modern empires like the Spanish, French and British, in which the home administration did not collapse—a notable difference from the empires of the past. Writing in 1988, of course, he was just a little early to include the collapse of the Soviet Union.

He finishes this chapter with a look at the characteristics of collapsed societies. As one would expect, with collapse comes a breakdown of authority and central control and the services provided at that level. The populace is left to get by as best they can and organize themselves at the local level, with much less long distance trade and communication. There is typically a marked rapid reduction in population size and density, because people flee and/or the death rate comes to exceed the birth rate. One might assume that the death rate increases, but if we look at the case of the Soviet Union, it is interesting that the birth rate actually went down significantly during the collapse.

Given all this, it is small wonder that collapse is feared by so many people today. Whether collapse is universally a catastrophe is considered in the concluding chapter of the book.

The second chapter considers the nature of complex societies. What follows is my impression of what Tainter has to say; apologies to him if I have misunderstood or added much that is my own.

At one end of the spectrum we have hunter gatherer societies, at the other our modern global industrial society.

The small hunter gatherer band is the archetypal un-complex society. This is not to say that considerable effort isn't spent in organizing and co-ordinating the endeavours of its members, but rather that at this scale it can all be done with nothing more than native human abilities. No administrative organization or written records, for instance.

Tainter describes several theories of how more complex societies got started. The one that makes the most sense to me is "Managerial": as we move to larger and more technologically advanced societies, the number of people involved and the differentiated and interconnected nature of their activities is such that an administrative organization must created in order to manage the whole thing.

There are two others that are referenced latter in sections of the book that I want to quote, so I'll include them here, conflict theory and integrationist theory:

In essence, conflict theory asserts that the state emerged out of the needs and desires of individuals and subgroups of a society. The state, in this view, is based on divided interests, on domination and exploitation, on coercion, and is primarily a stage for power struggles.

Integrations theories suggest that complexity and stratification arose because of stresses impinging on human populations and were positive responses to those stresses. Complexity then serves population-wide needs, rather than responding to the selfish ambitions of a few.

It seems that leadership is essential in complex societies and establishing the legitimacy of that leadership is a significant challenge. People accept leadership because they see that it brings them advantages or because they are coerced into doing so. Both strategies are expensive and this is the basis states and politics.

The third chapter looks at the study of collapse and the various factors to which scholars have attributed collapse. Tainter looks at 11 major explanations of collapse and finds them wanting in a variety of ways, offering the following summary at the end of the chapter:

  1. Resource depletion. Dealing with resource depletion is a common activity of complex societies and may be one of the thing that they do best. Where this is not the case, research has to focus on the characteristics of the society that prevent an appropriate response, rather than the exclusively on the characteristics of the depleted resource.
  2. New resources. This theme has some attraction of integration theorists, but none to conflict theorists. Its usefulness is mainly restricted to simpler societies.
  3. Catastrophes. Complex societies regularly provide for catastrophes, and routinely experience them without collapsing. If the society cannot absorb a catastrophe, then in many cases characteristics of the society will be of greater interest, obviating the catastrophe explanation.
  4. Insufficient response to circumstances. The assumptions made in this theme about the nature of complex societies—that they are inherently fragile, or static, or incapable of shifting direction—simply cannot be supported. Where complex societies may display such characteristics, that is a matter to be explained.
  5. Other complex societies. Major cases, such as the Roman one, cannot be accounted for by this theme. Conflict between states more often leads to cycles of expansion and contraction than to collapse.
  6. Intruders. The overthrow of a dominant state by a weaker one is an event to be explained, not an explanation in itself. Empirically, intruders are often difficult to detect archeologically where they have be postulated. It is difficult to understand why barbarians would destroy a civilization if it was worth invading in the first place.
  7. Conflict/contradictions/mismanagement. The capacity to control labor and allocate resources is intrinsic and necessary in complex societies. Collapse cannot be explained by factors so vital to survival, at least not without raising many more questions than are answered. Elite mismanagement and self-aggrandizement, to the extent that these are detrimental to the survival of a society are matters to be explained. Exploitation and misadministrations are normal, regular aspects of complex societies, and by themselves cannot account for... collapse. Peasants rarely revolt except when allied with other social strata, and their rebellions are not typically aimed at collapse.
  8. Social dysfunction. These explanations offer neither sources of strain nor causal mechanisms that can be analyzed in any objective way.
  9. Mystical. Mystical explanations fail totally to account scientifically for collapse. They are crippled by reliance on a biological growth analogy, by value judgments and by reference to intangibles.
  10. Chance concatenation of events. This theme provides no basis for generalization. Collapse is not well explained by reference to random factors.
  11. Economic explanations. These are structurally and logically superior to the others, at least as these others have been formulated to date. They identify characteristics of societies that make them liable to collapse, specify controlling mechanisms and indicate causal chains between controlling mechanisms and observed outcome. While economic explanations are not universally accepted in the social and historical sciences, such scenarios remedy the logical deficiencies of the other approaches. Existing economic models often suffer from incomplete forays into political and social explanations, but this is not an intrinsic flaw. The major drawback to economic explanations, for present purposes, is to develop an explanatory framework that is globally applicable.
With the exception of mystical explanations, which are without scientific merit, none of these explanatory themes fails entirely. Indeed, the economic theme comes close to success—in logic, if not in specifics—but does not go quite far enough. Except for mystical explanations, these approaches are not necessarily wrong or misguided. They are simply inadequate as presently formulated.... A general explanation of collapse should be able to take what is best in these themes and incorporate it. It should provide a framework under which these explanatory themes can be subsumed, so that one can account for what is worthwhile in each. A general explanation should make these themes clearer in application than each would be standing alone.

In Chapter 4, Tainter develops that explanation.

It will be no surprise to my readers that this explanation involves energy. While Tainter doesn't mention it, the ideas of "dissipative systems" and "dissipative structures" are useful here. Wikipedia defines a dissipative system as a thermodynamically open system which is operating out of, and often far from, thermodynamic equilibrium in an environment with which it exchanges energy and matter. It goes on to define a dissipative structure as a dissipative system that has a dynamical régime that is in some sense in a reproducible steady state.

Living organisms (including human beings) and ecologies are dissipative structures. So are human societies. All these structures maintain themselves by taking in energy and giving off waste heat. That "reproducible steady state" is not a matter of equilibrium, indeed it is definitely not is an sort of "balance", and if the supply of energy falls below the appropriate level, that state cannot be maintained and death and dissolution follow. On the other hand if a surplus of energy is available, these systems can grow and become more complex.

Indeed, my definition of politics is "the means by which a society decides what it will do with surplus energy".

Hunter gatherers, when faced with a windfall of surplus energy (usually in the form of food), often simply choose to accumulate a larger reserve of body fat, enjoy more leisure time and make sure more of their offspring survived to adulthood. This lifestyle worked very well for a couple of million years, during which time we spread to all the continents except Antarctica, and began to fill them up.

It seems to me that the problem which initially (and often since then) drove the adoption of complexity was this growing human population and the need for an increasing the food supply. It started out with the simplest forms of agriculture, which required more work and more organization than hunting and gathering, but returned a higher and more predictable yield of food on a smaller area of land. Remember, before the invention of heat engines, societies ran largely on muscle power, so food was the most important form of energy.

The important things that Tainter would have us understand is that each additional layer of complexity requires an additional amount of energy to support it. And further, that each additional investments in complexity give us lower returns, eventually reaching the point where more complexity actually makes things worse. This is called the law of "decreasing marginal productivity" and it is illustrated in this diagram, Tainter's Figure 19.

The horizontal axis represents the "Level of Complexity" and the vertical axis represents the "Benefits of Complexity". The shape of the curve gives us "decreasing marginal productivity." From the origin up to point B1C1 increases in the level of complexity yield very favorable increases in benefits. After B1C1, further investments in complexity still increase benefits, but by a smaller amount. For a similar investment today we get less improvement in benefits than we did yesterday and tomorrow is even worse. At point B2C2 additional complexity brings no benefits and at great deal is being spent just to maintain the status quo. If we go further to the right, adding complexity actually makes things worse. Note than at point B1C3, the benefits of all that added complexity are just the same as at the much less complex point B1C1. A move back to point B1C1 would be beneficial if it could be done without causing too much damage. The difficulty is that the very fabric of society is based on complexity, and at point B1C3, quite a lot of complexity. To switch back to point B1C1, much of that complexity would have to be dismantled or abandoned and replaced with something simpler—a process which, by definition, may be indistinguishable from collapse.

Human societies are problem solving organizations and our first choice as a problem solving technique is to add complexity. This often comes in the form of technology, which never makes things simpler, even though the added complexity may be hidden somewhere out of sight in the organization.

The horizontal axis of Fig. 19 can also be seen as representing energy used to maintain complexity, and the vertical axis as the energy gained by adding complexity. Initially, adding complexity yields enough extra energy to support that new level of complexity and enough surplus to support using further complexity to solve further problems.

Nothing succeeds like success and human societies tend to become like Christopher Robin's friend Tigger, who could only climb up trees. To get him down, drastic action was required (some help from the narrator, rotating the book so poor Tigger could step off the tree onto the text and from there to the ground). Human societies often get caught in "progress traps" and the only solutions they can think of just make things worse. Problems arise, and eventually things get to the point where adding complexity doesn't solve them. Collapse often ensues.

This can be avoided by "extending the tree"—adding technologies that use energy more efficiently or by acquiring additional sources of energy. Traditionally, this was done by taking over more land to grow more crops. In the modern world, fossil fuels have provided a huge energy subsidy which has enabled our society to become ever more complex. But as Figure 20 shows, this offers only a temporary reprieve. As we move further to the right on the curve, we eventually reach another area of decreasing returns and have to confront once again the possibility of collapse.

In any case, societies which find themselves on a downward sloping part of the curve are very vulnerable to collapse.

First, as the marginal return on investment in complexity declines, a society invests every more heavily in a strategy that yields proportionally less. Excess productive capacity and accumulated surpluses may be allocated to current operating needs. When major stress surges (major adversities) arise there is little or no reserve with which they may be countered. Stress surges must be dealt with out of the current operating budget. This often proves ineffectual. Where it does not, the society may be economically weakened and made more vulnerable to the next crisis....

Secondly, declining marginal returns make complexity an overall less attractive strategy, so that parts of a society perceive increasing advantage to a policy of separation or disintegration. When the marginal cost of investment in complexity becomes noticeably too high, various segments increase passive or active resistance, or overtly attempt to break away.

Tainter wraps up Chapter 4 by commenting that historically, the need for more energy was frequently met by conquering more land and this could provide only a temporary respite from declining marginal productivity. This tendency in particular resulted in a situation where collapse, when it did occur, affected a wider territorial sphere in a more devastating manner than might have otherwise been the case.

Next time, I'll wrap up this book review, covering Chapters 5 and 6 of The Collapse of Complex Societies.