A physicist or a chemist speaking of the achievements of his science would probably sound more self-assured than an economist, and he would have a right to. Economists labor under a series of handicaps from which the exact sciences are usually free. These handicaps are responsible for the frequent uncertainty of our knowledge, and for the unreliability of our forecasts. Let us note the principal ones.
Economists arc handicapped, in the first place, because economic reality is complex and hard to come to grips with. What with billions of people in the world, hundreds of thousands of commodities and prices, trillions of purchases and sales during the year, the man who advised, "Get your facts first," was handing out no small order. Statisticians and theorists have found means to summarize and simplify this complexity. But much valuable information gets lost in the simplifying shuffle.
Even more important, economists cannot experiment. Biologists can set up a colony of mice and put this society through its paces under varying controlled conditions. Humanity's ardor for more precise economic knowledge has not been great enough, so far, to supply economists with similar facilities. Economists can, of course, gather experience, including such experience as may be derived from measures taken at their advice. The early New Deal undertook, on the advice of economists, a series of measures that in retrospect look like experiments. We learned something from our ill-advised efforts to get out of the depression by devaluating the dollar and regulating the economy through the N.R.A. —how not to do it. But economists do not deliberately experiment with the fate of a nation, and from the scientist's point of view such experience lacks the decisive ingredient of a real experiment—control of the environment, and repetition.
Statistical tools have been forged that enable economists to trace cause and effect even under the haphazard experimental conditions offered by nature. But they are a poor substitute for the order and simplicity with which many other sciences can arrange their subject-matter. In economics, nothing is certain, anything is possible, and everything depends on everything else. The plain difficulty of understanding what goes on in the economy is the first big handicap faced by economists.
This insufficient understanding naturally adds to the complications of forecasting. If we have a hard time keeping up with people's present
actions, how can we appraise their intentions for the future? Often people themselves do not know what they intend to do and, if they do, they often change their minds. As a matter of fact, economists may be able to predict people's behavior better than people themselves can. The action of a single individual is quite uncertain, but the behavior of large groups follows the laws of probability. Insurance companies base their calculations upon this rule of large numbers.
Two factors, however, often frustrate economists who would like to avail themselves of the laws of probability. The first is the prevalence of trends and fashions among consumers and business men. Like an epidemic that upsets the mortality tables, sudden trends and fashions upset the economist's probability calculations. The second is the dominant role of a few important decision makers—leaders in politics, business and labor. Their actions cannot be forecast by the laws of large numbers, men in this elite category not being numerous. Yet their decisions, essentially unpredictable, influence the course of events.
It is worth noting also that economic forecasting, to be successful, has to be extraordinarily accurate. Whether business will turn up or down, for instance, depends on a very slim margin. If total demand exceeds total supply by a couple of billions out of about $430 billion of gross national product, business will expand. If demand falls short by a similar sum, business will contract. Economists can easily predict that total demand will be approximately $430 billion, but that is not good enough. A difference of two billion either way, less than half of 1 per cent, means the difference between up and down. This applies to most forecasts in economics.
After all this, it should be obvious why economists do not always agree. It might be more to the point, however, to emphasize that economists do agree on a great many things that make up the body of their science. Being agreed upon and accepted, these points are just not discussed much. Today a large measure of agreement exists even in formerly controversial fields like the general principles of anti-recession policy, as recent Congressional hearings have demonstrated. There the argument has been mainly over technical details and timing.