THE GREATEST ADVANCES IN SOCIAL WELFARE MEASUREMENT HAVE BEEN IN THE FIELD OF ECONOMICS, YET EVEN THERE THE COMMON MEASURES ARE UNDER ATTACK BECAUSE THEY DO NOT TRULY MEASURE WHAT THEY SEEM TO. THE INCOME OF THE POOR IS CONSISTENTLY UNDERESTIMATED BECAUSE COMMON REPORTING RATES DO NOT TAKE INTO ACCOUNT FREE MEDICAL CARE, FOOD STAMPS, OR OTHER BENEFITS. YET SUCH BENEFITS ARE THE MOST RAPIDLY GROWING ECONOMIC COMPONENT IN AMERICAN SOCIETY. CRIME RATES ARE MEASURED IN TERMS OF CRIMES ACTUALLY REPORTED TO THE FEDERAL BUREAU OF INVESTIGATION AND THIS IS CONSIDERED ANALOGOUS TO MEASURING UNEMPLOYMENT BY COUNTING THE NUMBER OF PEOPLE WHO VISIT A STATE UNEMPLOYMENT SERVICE. (BOTH MEASURES ARE CONSIDERED INVALID.) THE PROBLEMS OF UNDERREPORTING ARE GREAT. MEASURING ATTITUDES IS EQUALLY DIFFICULT--AS DE TOCQUEVILLE POINTED OUT, THE TOLERATION OF BAD CONDITIONS DECREASES AS CONDITIONS IMPROVE. LIMITATIONS IN THE ABILITY TO MEASURE SOCIAL WELFARE CERTAINLY HAVE IMPLICATIONS FOR DEVELOPING EFFECTIVENESS MEASURES FOR FEDERAL PROGRAMS; HOWEVER PROBLEMS IN DEVELOPING OVERALL MEASURES NEED NOT PREVENT THE ACHIEVEMENT OF SMALL ADVANCES IN THE ABILITY TO DIAGNOSE AND PRESCRIBE WHICH, VIEWED CUMULATIVELY, MARK PROGRESS IN IMPROVING THE QUALITY OF LIFE. (GLR)