IS ECONOMICS A SCIENCE? By Henry C. Wallich Economists are having a trying time these days. "Why can't you say how long this recession will go on? Why can't you agree about what ought to be done? What good is your science if people don't know whether they can rely on your predictions and prescriptions?" Economists have heard this sort of thing before and acutely aware that they have much to be modest about. They remember Bernard Shaw's observation that if the economists of this world were laid end to end they wouldn't reach a conclusion. They have heard angry young men within their own ranks question economic science on the ground that where it is really scientific it doesn't have much to do with economics, and where it is economics it isn't scientific. They may long for the golden days of economics a hundred and more years ago, when the science was held in awe and the views expressed by economists were received with appropriate respect.
If economists were to take time out to reply to today's charges they might say something like this: It is, of course, quite true that eco-nomics is not an exact science. But economics can claim a number of important discoveries that have improved our understanding and our economic performance. If economists cannot make reliable forecasts nor always arrive at universally agreed recommendations, the failure is attributable to difficulties of a kind that sciences like physics and biology do not have to face.
Note for a moment such progress and accomplishments as can be found in economics. The field is wide—economics is what economists do, it has rightly been said, and economists nowadays do many things that Adam Smith never dreamed of. One can perhaps divide economists into three groups in accordance with their activities, doing about equal violence to each by this arbitrary classification. One group comprises the finders and analysts of facts the second the theorists, the third the administrators and policy makers. Many economists, of course, qualify as members of more than one group.
The fact-finders have done a very solid job. Time was when economic statistics were in the main a by-product of some primitive counting of noses and pennies that governments conducted for fiscal purposes.
Foreign trade statistics, for instance, have always been notably complete because goods flow into and out of a country through a few narrow channels and are duly tagged by the customs people. Today we have highly elaborate information in many other areas as well.
Consider, for instance, the national income statistics, which show the total of all incomes received in the country. We know the origin of the national income—how much is contributed by_ industry, agriculture and services. We know its uses—how much goes for consumption, how much is saved and invested. We know its distribution—how much of the total goes to labor, to capital, to professional people and others. And many of these data come to us at monthly intervals, with a lag of barely one month.
Surely it is worth a good deal in our present recession to know just where we stand. Without data like these, people with conflicting axes to grind would no doubt inform us, on the same day, that the country was facing bankruptcy and that there was no recession at all. Any effort to make economics a more exact science must begin with statistics.
Another feather in the fact-finder's cap is the development of data on plans and intentions. Consumer spending plans for durable goods are regularly surveyed now. Business plans for expenditures on new plant and equipment are reported for two and three years ahead. What the factfinders don't find it easy to anticipate is how people are going to change their minds. But, even without that, the intentions data are helpful in forecasting.
Meanwhile, very elaborate methods for analyzing economic data have been developed by mathematical statisticians. Gone are the days when statistics just meant collecting figures and adding them up. An advanced mathematical "model" of the economy is an elaborate assembly of formulas that can tax the powers of an electronic computer. The practical results, to be sure, are still often inconclusive or worse. The analysts are the first to admit that their technique is in its infancy and that they have more to learn than to teach.