In 1950 appeared the first edition of Oskar Morgenstern's famous book,
The Accuracy of Economic Observations, Nearly half a century later it
is timely to return to Morgenstern's diagnosis and to contemplate his
therapeutic recommendations. Morgenstern's vision can and should info
rm the consideration of the topic today because of the continued valid
ity of many of his findings, His work still provides stimuli for study
ing the general problems of measurement, the varying requirements for
accuracy, the issues of aggregate macroeconomic measures, and the pros
pects for economic and social measurement, This is so even if some of
the bleaker assessments by Morgenstern, notwithstanding their technica
l merits, provide little or no practical guidance for statistical acti
vities, In this context it is enlightening to recall the different pra
ctical attitudes adopted by Keynes and by some of his contemporaries i
n Germany regarding theoretical difficulties with aggregate macroecono
mic data.