THE RELATION OF COMPUTER MODELS TO ECONOMIC FORECASTING WAS THE SUBJECT OF AN OPEN DISCUSSION SPONSERED BY THE COMMITTEE ON ECONOMIC FORECASTING OF THE DEPARTMENT OF ECONOMICS, FINANCE AND ADMINISTRATION. ALL PANELISTS RECOGNIZED THE COMPUTERS VALUE IN TERMS OF SPEED, BOOK- KEEPING ABILITY, QUICK TESTING OF ALTERNATIVES AND THE ABILITY TO GIVE MATHEMATICAL OR STATISTICAL FORM TO LOGICAL IDEAS AS A GUIDE FOR JUDGEMENTS AND POLICIES. HOWEVER, MANY WERE AGAINST COMPLETE ACCEPTANCE OF A MECHANICAL FRAMEWORK OR TOO MUCH DEPENDENCE ON THE INFALLIBILITY OF ARBITRARY PROGRAMS. TOO LITTLE ROOM FOR SUBJECTIVE INPUTS WAS THOUGHT TO BE A MAJOR DISADVANTAGE. FOR EXAMPLE, COMPUTER MODELS ARE INSENSITIVE TO PUBLIC VALUE JUDGEMENTS. USEFULLNESS OF COMPUTER MODELS WAS AGREED TO BE DEPENDENT UPON ADEQUATELY WEIGHTED INPUT DATA, AND CRITICAL ANALYSIS OF COMPUTER TAPES.