Normal Accidents and Macro-Systems

What does the Chernobyl accident have in common with the threat of Global Warming? Clearly, both involve failures of modern technology but in other respects they seem quite different. At Chernobyl a poorly designed reactor was sent into meltdown due to human error. If the designers or operators could have foreseen the outcome they would certainly have done things differently and better. Global Warming is another matter entirely. It involves not the design of a single technology but rather the design of most technology. Whether operators err or not is irrelevant. In either case they contribute to the “accident” or what more accurately would be described as an unintended side-effect.

Yet I want to show that these two cases have something important in common.

The Chernobyl case is an example of what Charles Perrow has called a “normal accident.” Normal accidents result from two aspects of modern technological design. First, many technologies depend on an enormous number of interactions between their parts, so many that their very complexity makes them unpredictable. And second, the tight coupling of parts means that when one is affected the consequences propagate quickly throughout the system.

This is certainly true of a nuclear power plant which is why so much care must go into design, operation, and training of personnel. Thousands of interconnected parts react to the movement of a single switch or a small change in temperature within the reactor core. This is not the kind of technical device human beings have been accustomed to creating and using since they evolved tens of thousands of years ago.

Compared to traditional technical devices, complex modern technologies belong to a new and unnatural world. Consider the tools of an old fashioned farm. These tools are no doubt products of genius but of a very different kind of genius than our modern technology. Each tool is relatively simple and understandable. The farmer expects to be able to repair them all many times in the course of their useful life. And the breakdown of one tool usually has no effect on the others which can still be used while the broken tool is under repair. In sum, the farmer’s tools cannot get out of control and do something untoward that destroys the whole farm or kills the farmer. Of course accidents happen on farms all the time, but they are not “normal accidents.” Rather, they are due to failures of single components or alertness and are usually easily understood.

In so far as modern life is built around complex modern technologies, we have to get used to normal accidents and attempt to foresee and prevent as many of them as possible. This task requires a new and far more cautious mentality that is not easily learned even by people with sophisticated training. Indeed, well trained individuals may have excessive confidence in their own ability to control complex systems, an error of perspective that led to the failures of the American space shuttles. It is difficult to overcome the habits of millennia during which behavior we would now consider sloppy and dangerous was just fine.

Turning now to Global warming, we find similarities and differences. While normal accidents are unpredictable, we were warned about Global Warming in 1898, over a century ago, by the Swedish scientist Svante Ahrrenius. And yet practically nothing has been done to fend off the impending catastrophe. Why not?

The problem results in part from the disciplinary boundaries that separate the sciences and engineering specialties. During the last century of rapid technical progress, there was no reason for engineers developing automobiles and power plants to think about the environmental consequences of their work. The very specialization that made them capable of executing the tasks assigned them blinded them to whole regions of the real world in which their products would have to function. These disciplinary boundaries and specializations are inherited from an earlier era when science had relatively few technological applications. The difference between chemistry, physics, and medicine corresponded to properties of matter human beings could not yet bring into relation in subtle ways with complex long term effects.

We now live in a technical environment in which all these connections are made. The isolation of the sciences, medicine, and engineering is breaking down. New disciplines such as ecology begin to pull together a far more concrete and realistic image of nature and of our place in it. Ahrrenius’s prediction is now charted by sophisticated computer models based on knowledge from many disciplines.

But even this scientific advance is insufficient to get much action. The forces leading to Global Warming are social and economic as well as technical. It is one thing for an engineer to change a component, quite another to redesign a socio-economic system. And yet modern society and the modern economy have become intricate machines, not so very different from the complex systems that suffer normal accidents. Automobiles, roads, oil supply and refining, rubber supply and tire production, job classifications and management techniques, factory design and electrification, suburban lifestyles and urban design, the situating and growth of cities, all are intertwined in such complex ways that sorting out the connections is virtually impossible. This is the world of what the French sociologist Alain Gras calls “macro-systems,” gigantic socio-technical systems with roots and branches throughout the social world.

It is true that the couplings are looser than in a nuclear reactor but control is also far weaker since so many different individuals and agencies are involved. The fact of complexity makes it hard to change the system even once problems are identified. The environmentalist must convince millions, nay, billions of people to change their way of life in order to introduce a environmentally sound technology! The issues are political and call for education and mobilization of opinion on a large scale.

The result of this comparison is rather discouraging. We are confronted with the probability of normal accidents where we cannot make predictions and with the probability of inaction in the face of disaster where we can make predictions. Complexities of different types are the problem in both cases. The solution, to the extent that there is one, is caution and education. But the historical precedents suggest a further obstacle.

The modern economy was created by capitalist entrepreneurs. Caution was not their strongest virtue. Rather, they were a daring breed who introduced rapid, wholesale changes with little thought for the long term future. We are the heirs of a world built by recklessness, not caution. As for education, it is increasingly restructured to service the labor market. Technical specializations are privileged over the kind of wide-ranging learning that would prepare young people to make better decisions about the future. Meanwhile, far more influential than education, the mass media project images of the good life that are environmentally unsound and unsustainable. It is no wonder that many observers today despair of the future.

But I think there is considerable room for hope despite all the bad news. Two of the most destabilizing trends in the environment appear to be self-regulating. The demographic transition is engaged on a global scale, slowing population growth which appeared to be completely out of control 20 years ago. Economic forces at work in the domain of energy will likely force drastic changes in modern methods of production and transport over the next decades, greatly increasing the energy efficiency of global technology.

Still, there is no guarantee that the necessary adaptations will be beneficial for the environment. The risk is that those changes will in turn expose us to ever more “normal accidents” as complex technologies such as nuclear power multiply to maintain the level of energy consumption.

Better responses are possible, but for them to win out, governments must help the many victims of progress today in addressing the issues and bringing them to the attention of the public. Only in this way can the political forces be mobilized to introduce a new model of progress based on caution and understanding rather than the recklessness and ignorance of the past.