The Effects of Nuclear War

UNCERTAINTIES

There are enormous uncertainties and imponderable involved in any effort to assess the effects of a nuclear war, and an effort to look at the entire range of effects compounds them. Many of these uncertainties are obvious ones: if the course of a snowstorm cannot be predicted 1 day ahead in peacetime, one must certainly be cautious about predictions of the pattern of radioactive fallout on some unknown future day. Similar complexities exist for human institutions: there is great difficulty in predicting the peacetime course of the U.S. economy, and predicting its course after a nuclear war is a good deal more difficult. This study highlights the importance of three categories of uncertainties:

• Uncertainties in calculations of deaths and of direct economic damage resulting from the need to make assumptions about matters such as time of day, time of year, wind, weather, size of bombs, exact location of the detonations, location of people, availability and quality of sheltering, etc.
• Effects that would surely take place, but whose magnitude cannot be calculated. These include the effects of fires, the shortfalls in medical care and housing, the extent to which economic and social disruption would magnify the effects of direct economic damage, the extent of bottlenecks and synergistic effects, the extent of disease, etc.
• Effects that are possible, but whose likelihood is as incalculable as their magnitude. These include the possibility of a long downward economic spiral before viability is attained, the possibility of political disintegration (anarchy or regionalization), the possibility of major epidemics, and the possibility of irreversible ecological changes.

One major problem in making calculations is to know where the people will be at the moment when the bombs explode. Calculations for the United States are generally based on the 1970 census, but it should be borne in mind that the census data describes where people’s homes are, and there is never a moment when everybody in the United States is at home at the same time. If an attack took place during a working day, casualties might well be higher since people would be concentrated in factories and offices (which are more likely to be targets) rather than dispersed in suburbs. For the case of the Soviet population, the same assumption is made that people are at home, but the inaccuracies are compounded by the unavailability of detailed information about just where the Soviet rural population lives. The various calculations that were used made varying, though not unreasonable assumptions about population location.

A second uncertainty in calculations has to do with the degree of protection available. There is no good answer to the question: “Would people use the best available shelter against blast and fallout?” It seems unreasonable to suppose that shelters would not be used, and equally unreasonable to assume that at a moment of crisis all available resources would be put to rational use, (It has been pointed out that if plans worked, people behaved rationally, and machinery were adequately maintained, there would be no peacetime deaths from traffic accidents. ) The Defense Civil Preparedness Agency has concluded from public opinion surveys that in a period of severe international crisis about 10 percent of all Americans would leave their homes and move to a “safer” place (spontaneous evacuation); more reliable estimates are probably impossible, but it could make a substantial difference to the casualty figures.

A third uncertainty is the weather at the time of the attack at the various places where bombs explode. The local wind conditions, and especially the amount of moisture in the air, may make an enormous difference in the number and spread of fires. Wind conditions over a wider area determine the extent and location of fallout contamination. The time of year has a decisive effect on the damage that fallout does to agriculture–while an attack in January might be expected to do only indirect damage (destroying farm machinery or the fuel to run it), fallout when plants are young can kill them, and fallout just before harvest time would probably make it unsafe to get the harvest in. The time of year also has direct effects on population death — the attack in the dead of winter, which might not directly damage agriculture, may lead to greater deaths from fallout radiation (because of the difficulty of improvising fallout protection by moving frozen dirt) and from cold and exposure.

The question of how rapid and efficient economic recovery would be— or indeed whether a genuine recovery would be possible at all — raises questions that seem to be beyond calculation. It is possible to calculate direct economic damage by making assumptions about the size and exact location of bomb explosions, and the hardness of economic assets; however, such calculations cannot address the issues of bottlenecks and of synergy. Bottlenecks would occur if a key product that was essential for many other manufacturing processes could no longer be produced, or (for the case of a large attack) if an entire industrial sector were wiped out. In either case, the economic loss would greatly exceed the peacetime value of the factories that were actually destroyed. There does not appear to be any reliable way of calculating the likelihood or extent of bottlenecks because economic input/ output models do not address the possibility or cost of substitutions across sectors. Apart from the creation of bottlenecks, there could be synergistic effects: for example, the fire that cannot be controlled because the blast destroyed fire stations, as actually happened at Hiroshima. Here, too, there is no reliable way to estimate the likelihood of such effects: would radiation deaths of birds and the destruction of insecticide factories have a synergistic effect? Another uncertainty is the possibility of organizational bottlenecks. In the most obvious instance, it would make an enormous difference whether the President of the United States survived. Housing, defined as a place where a productive worker lives as distinct from shelter for refugees, is another area of uncertainty. Minimal housing is essential if production is to be restored, and it takes time to rebuild it if the existing housing stock is destroyed or is beyond commuting range of the surviving (or repaired) workplaces. It should be noted that the United States has a much larger and more dispersed housing stock than does the Soviet Union, but that American workers have higher minimum standards.

There is a final area of uncertainty that this study does not even address, but which could be of very great importance. Actual nuclear attacks, unlike those in this study, would not take place in a vacuum. There would be a series of events that would lead up to the attack, and these events could markedly change both the physical and the psychological vulnerability of a population to a nuclear attack. Even more critical would be the events after the attack. Assuming that the war ends promptly, the terms on which it ends could greatly affect both the economic condition and the state of mind of the population. The way in which other countries are affected could determine whether the outside world is a source of help or of further danger. The postattack military situation (and nothing in this study addresses the effects of nuclear attacks on military power) could not only determine the attitude of other countries, but also whether limited surviving resources are put to military or to civilian use.

Moreover, the analyses in this study all assume that the war would end after the hypothetical attack. This assumption simplifies analysis, but it might not prove to be the case. How much worse would the situation of the survivors be if, just as they were attempting to restore some kind of economy following a massive attack, a few additional weapons destroyed the new centers of population and of government?

---