… is from page 305 of the 1963 second edition of Oskar Morgenstern’s great 1950 volume, On the Accuracy of Economic Observations (footnote deleted):
However, no amount of organization can replace the work of the economist in teaching and research. Students have to be brought up in an atmosphere of healthy distrust of “the facts” put before them. They must also learn how terribly hard it is to get good data and to find out what really is “a fact.”
DBx: Note that Morgenstern does not argue that facts are irrelevant or unable to be discovered. Nor does he argue that one person’s ‘facts’ are as good (or as bad) as any other person’s facts. Instead, his argument, in part, is that, especially in the social sciences, discovering “a fact” is typically far more difficult than most people, including many economists, suppose. For example, a number on a dot-gov webpage might in fact be a fact, as we understand that term, but it also might be a pseudo-fact: something that appears to be objective and true but really does not measure what those who construct and report it present it as measuring. (For some specific examples, see again this new video by Russ Roberts.)
Morgenstern’s argument also includes at least two other points. One is the recognition that not all ‘true’ facts are equally relevant or revealing; indeed, many ‘true’ facts have no relevance whatsoever. An example of an utterly irrelevant ‘true’ fact is the U.S. trade deficit with China. This fact seems to many people to have great relevance, but it has none. Or, to put the point differently, knowledge of this ‘fact’ is no more useful for assessing an economy’s performance and for guiding economic policy than is, say, accurate knowledge of the number of birds currently nesting in trees along the roadside for guiding you on your drive to your destination. There is such a number (assuming we agree on what is meant by “birds,” “nesting,” “trees,” and “along the roadside”!), but you would find knowledge of that number to be of no use for your purposes.
Second is the recognition that facts are meaningless without a theory to categorize and assess them. Bluntly, every ‘fact’ implies a theory used to give that fact meaning and relevance. The theory might be ‘good’ or it might be bad. But no one – absolutely no one – is driven only by “the data” or only by “the facts,” if for no reason other than that “the data” or “the facts” have no meaning to a human being independently of the theory that the observer of the data and the facts uses, perhaps unawares, to make sense of them.
