SOCIAL statistics have become a national obsession. Optimistic that problems can be solved, Americans want to know the answer to ''How Are We Doing?'' in virtually every area of social and political life.
Responding to this demand, researchers have devised quantitative measures of just about everything. In the economic sphere we track every twist and turn of the gross domestic product (GDP), family income, poverty, inflation, productivity, and much more. On matters environmental, we seek to gauge risks involved in using an enormous array of chemicals, from agricultural pest control to household aerosols.
The potential of assorted ''megatrends and tendencies'' -- America's competitive position in the world economy, global warming, world population growth, energy supply and consumption -- are insistently measured. If effort, amounts spent, and volume of results are proper indicators in this quest for social measurement, the end product is impressive.
Unfortunately, they are fundamentally irrelevant to the main issue. What we need to know, of course, is how good our measures are. How accurately do they describe the social conditions they claim to describe?
In the June/July issue of Public Perspective Magazine, my colleagues and I review a series of just-published books and articles that document how great the error is in many widely used indicators. Measurement is often extremely complex. In addition, as Richard C. Lewontin, a distinguished population geneticist at Harvard University, observed in a recent essay, efforts at measuring social behavior and performance confront the inevitable fact that ''human societies are made by self-conscious organisms.'' We care about our social performance and try to influence it.