On the Accuracy of Economic Observations

Tus BooK addresses itself to both the general reader and the professional economist. The former will see that decisions made in business and in public service are based on data that are known with much less certainty than generally assumed by the public or the government. The second will discover that even most widely accepted figures frequently have error components of unexpected magnitude, and consequently cast doubt on many currently accepted analyses in economics. We all shall have to go through a long and painful process of adjusting to the fact of error. Although the natural sciences—sometimes called the “exact” sciences—haye been concerned with the ac- curacy of measurements and observations from their earliest beginnings, a great crisis was nevertheless suffered by them when it became clear that absolute precision and certainty of important kinds of observations were impossible to achieve in principle. The situation in the social sciences may not be sim- pler, but even if it should be, there exists no comparable tradi- tion of exploring errors and evaluating their influence upon economic science and its application. However, when the true conditions are realized, there will evolve a more powerful and realistic theory. We must carefully distinguish between what we think we know and what we really do and can know. The second edition of this work has been almost completely rewritten. The first edition, published in 1950, went out of print in 1952 but the republication was postponed time and again un- til now. Other work has been responsible for the delay, not- withstanding the fact that the problems treated in the present work have always occupied a central role in my life. The temp- tation to let the book grow was therefore very strong and, de- spite efforts to resist, this volume is now more than twice as large as its predecessor. I hope it still strikes a good balance be- tween readability and thoroughness. Indeed, completeness is out of the question, since the number of economic data is limit- Econometric Research Program Princeton University April 1963 Oskar MORGENSTERN