One would be cloistered indeed not to have noticed that over the past year or so less than normal precipitation has characterized much of the continental United States.
Crops and water supplies are threatened. For example, 1980 yields for several crops (including corn and peanuts) were reduced by the heat wave and dry spell that baked the Midwest and South, and last winter's snowpack in the western mountains was one of the lightest ever recorded. And to the north and east of the Delaware Valley, restrictions on water use, imposed last fall and winter after one of the driest summers on record, remained in effect well into 1981.
These events so moved the editors of Newsweek that their February 23 cover depicted an outline of the forty eight contiguous states filled in their entirety with drought-parched earth. And inside, the feature story reflected the cover art. U.S. News and World Report devoted its June 29 cover story to water scarcity, and many other publications, from the revivified Life to Organic Gardening, also have given cover treatment to the issue. Thus, the "drought" of 1980-81 afforded the media promising material for analysis, as well as a basis for gloomy predictions.
Nowhere has this theme been more enthusiastically pursued than in the Northeast, particularly in New York, where the national newspapers and television networks have made much of regional and nationwide droughts. This could be considered harmless, except that under the dual stresses of actual precipitation shortfalls and hectoring newspaper and TV stories, water system managers and municipal and federal agency officials may commit themselves to unwise projects or policies.
To better understand the potential for distortion, it is useful to review what is known about the 1980 drought in the urbanized Northeast; how it was analyzed in the press, especially by The New York Times; and how that analysis reinforces traditional approaches that are biased toward expensive increments to water-system capacity.
The 1980 drought
Precipitation shortfalls, even when regionwide, vary in intensity so that no single set of measurements captures the regional experience. But some easily available measures show that the 1980-81 water shortage really was severe in the Northeast. For example, New York City's total precipitation for the three summer months of 1980 was 3.42 inches, the lowest ever recorded and, as significant, more than 20 percent lower than the second driest summer, that of 1916.
The summer of 1980 also was one of the hottest on record, increasing the demand for water for pools, commercial air conditioning, lawn irrigation, and fire hydrant sprinklers. The combination resulted in more rapid than normal reservoir drawdown (see figure 1). And as the dry spell continued into the fall, the reservoir levels kept falling, producing a percentage-of-capacity trace for the giant New York City system that approximated, until mid-February 1981, that of 1964 and 1965—the two worst years of the mid-1960s drought.
As reservoir levels fell in the fall and winter, one water system after another announced drought alerts, warnings, and emergency situations, and imposed restrictions on car washing; hosing down streets, driveways, and sidewalks; and operating ornamental fountains. To publicize restrictions, restaurants were asked to serve water only on request; and a great deal of advice was offered about showering, shaving, flushing, and dishwashing.
In some parts of northern New Jersey and southwestern Connecticut an attempt was made to enforce real rationing by limiting household connections to 50 gallons per capita per day, by increasing meter reading frequency, and by applying either criminal penalties or water surcharges to those who appeared to be in violation.
These measures gave one a sense of déjà vu, since they were the same as those imposed during the several drought years of the 1960s. Granted, some lessons had been learned from the California experience of the late 1970s about how far authorities could push people to change their habits in such sensitive areas as toilet flushing, but no water systems had installed remotely read meters or other technological innovations that would increase the feasibility of a truly effective rationing scheme.
Figure 1. Percentage of capacity eld in New York City reservoirs
"Drought" perspective
The drought of the 1960s, a much more severe event overall than the 1980-81 dry spell, came almost immediately after New York City's water system had completed its three Delaware watershed reservoirs. The system's "safe yield"—that quantity of water deliverable every day in all but 5 percent of years (based on what then was the worst drought on record)—was then about 1.55 billion gallons per day. With consumption averaging about 1.2 gallons per day, then-Commissioner D'Angelo might be forgiven his impetuous statement, made in January 1961, that the city would never again experience a water shortage. As the drought years followed each other, however, the existence of new reservoirs discouraged the media from blaming inadequate planning and insufficient investment for the shortage of water. It seemed clear, even to the Times, that this was a natural event so severe that it overwhelmed the most ambitious of plans and the most generous of investment programs.
In 1980-81 the city's physical capacity is the same as in the 1960s, but water withdrawals from it on an average day now are about 1.5 billion gallons, close to safe yield. Other systems in the region are in similar shape, whether because of environmental objections to the construction of new reservoirs or because investments in new sources were postponed as a result of budget constraints. (In northern New Jersey, two reservoirs had been built by the state after the 1960s drought, but they are not connected to the water systems that had problems in 1980-81.)
This left an irresistible opening for the media analysts. In February 1981, the Times stated that the need for water use restrictions has raised the question of why something was not done to prevent the shortages, especially when water resource experts were saying that even in times of drought there was plenty of water in the region for all essential uses." There followed some agitation for measures to prevent shortages, but media skeptics were not sanguine that citizen concern would survive the end of the drought. Whenever that came.
Some observations about droughts, shortages, and expansion of water system capacity.
The quotation from the Times captures the essence of a polar view of municipal water system planning: shortages are unacceptable, and capacity should always be large enough, relative to demand, to prevent them. This view is based on comprehensive misunderstanding, and attempts even to approximate its implications for investment are bound to be enormously expensive—perhaps even infeasible in the heavily populated Northeast.
Fundamentally, it is impossible to prevent shortages because, short of zero, there in no lower limit to possible precipitation over any given period. No matter how much storage capacity is provided, a natural event can make it impossible to meet all demands at whatever price normally is charged.
Second, shortages do not fit the characteristics implicit in the no-shortage position. Except in the kind of extreme situation that is experienced only very rarely and usually by systems without any significant storage, shortages do not arise out of an actual lack of water. Shortages are "created" by system managers who impose water-use restrictions long before their reservoirs go dry, precisely in order to prevent that from happening. This is done because their own intuition and the best available evidence say that damages are a steeply nonlinear function of shortage size. Thus, if there is overperiod storage in the system, and if it seems likely that in one period the available supply might be as much as 30 percent below demand, it often seems prudent to "save" a percent over each of three periods, rather than 30 percent in one period.
Given almost complete uncertainty about weather conditions for more than five days ahead, no one knows when it is necessary to reduce withdrawals by 30 percent. In retrospect, it often proves unnecessary to have imposed restrictions on use: it usually rains enough to have supported normal use. But not always—and in a multiyear drought, the system could court trouble by forecasting such a happy ending. Hence risk-averse managers create shortages, even though enough water may be physically available.
Table 1. Some Annual Drought Damage Estimates
Also, while shortages may be annoying and even psychologically disturbing to some people, they have quite modest economic effects until they become very large indeed. Pseudo shortages almost always begin by involving restrictions on such uses as car washing and pavement cleaning, surely nonessential tasks by any reasonable view. Even restrictions on watering lawns and shrubs cost consumers little unless the precipitation shortfall is severe enough to kill valuable trees or shrubs. Restrictions on commercial activity usually involve only nonrecycling systems and can easily be avoided by modest investments (some of which should perhaps have been made in any event).
Some data on drought losses from an RFF-funded study of the 1960s drought in Massachusetts, and a U.S. Corps of Engineers' study of the same event in York, Pennsylvania, produced estimates of losses as connected to reductions in water use (percentage "shortages"). See table 1.
In short, the appropriate response to pervasive uncertainty about future precipitation levels is planning to have shortages and not planning to prevent them. That is, municipal water systems should be designed so that, at the margin, the cost of increasing capacity is equal to the expected damages avoided by that addition. Because total expected damages usually can be reduced by intelligent and foresighted application of restrictions before there is a physical need for them, the expected damage estimates should be developed by costing the effects of programs of staged drought response, which include gradually tightened restrictions.
This is, in fact, the approach that has been taken by the Washington (D.C.) Suburban Sanitary Commission in deciding on its latest increment to capacity. For the sake of New York and New Jersey residents, one can hope that this is the lesson drawn from the 1980 drought, and not the one that has been portrayed by the news media.
The above article was written by Clifford S. Russell, director of RFF's Quality of the Environment Division.
A version of this article appeared in print in the October 1981 issue of Resources magazine.
