Due to the success of science in the 1800’s, people began to use the scientific method in their study of social sciences, such as economics. They construct hypotheses, go through history to gather data, and from this data they make conclusions and laws. I call this methodology malpractice. Milton Friedman was guilty of methodology malpractice. Friedman’s work, “A Monetary History of the United States”, displays this. While I find these types of works interesting, to make theories from historical data in social science is absurd.
Economics is the study of human action. Physics can tell us with precision where a cannonball will land if it launched at 45 degrees, with a velocity of 10m/s. Humans are not rocks or intimate objects. Individuals react very different to stimuli. A particular individual might react to the same stimuli different one year from now. The movie “The Matrix” points this out.
Let us say the money supply doubled in a particular ten year period in history and price inflation was 5%. Put aside the issue of the definition of money supply and inflation and how it is measured. There is no way it could be said the same thing will happen in the future. This historical information only tells us about a particular period of time. No laws can be deduced from this statistic.
Laws in social science are found by deductive logic, not be empirical data.
An entire field of economics called econometrics has been developed. It is taught in all classrooms. The Federal Reserve, other central bankers, governments, investment firms, etc. use this as a means of quantitative analysis. Econometrics contributes nothing to the study of economics. It is all nonsense.
One of Hayek’s greatest works was “The Counter Revolution of Science”. The entire book is dedicated to attacking scientism. Below, is what I consider, his most profound statements in the whole book.
Nothing is more instructive than to compare the nature of these statistical wholes, to which the same word “collective” is sometimes also applied, with that of the wholes or collectives with which we have to deal in the theoretical social sciences. The statistical study is concerned with the attributes of individuals, though not with attributes of particular individuals, but with attributes of which we know only that they are possessed by a certain quantitatively determined proportion of all the individuals in our “collective” or “population.” In order that any collection of individuals should form a true statistical collective it is even necessary that the attributes of the individuals whose frequency distribution we study should not be systematically connected or, at least, that in our selection of the individuals which form the “collective” we are not guided by any knowledge of such a connection. The “collectives” of statistics, on which we study the regularities produced by the “law of large numbers,” are thus emphatically not wholes in the sense in which we describe social structures as wholes. This is best seen from the fact that the properties of the “collectives” with statistics studies must remain unaffected if from the total of elements we select at random a certain part. Far from dealing with structures of relationships, statistics deliberately and systematically disregard the relationships between the individual elements. It is, to repeat, concerned with the properties of the elements of the “collective,” though not with the properties of particular elements, but with the frequency with which elements with certain properties occur among the total. And, what is more, it assumes that these properties are not systematically connected with the different ways in which the elements are related to each other.
The consequence of this is that in the statistical study of social phenomena the structures with which the theoretical social sciences are concerned actually disappear. Statistics may supply us with very interesting and important information about what is the raw material from which we have to reproduce these structures, but it can tell us nothing about these structures themselves. In some field this is immediately obvious as soon as it is stated. That the statistics of words can tell us nothing about the structure of a language will hardly be denied. But although the contrary is sometimes suggested, the same holds no less true of other systematically connected wholes such as, e.g., the price system. No statistical information about the elements can explain to us the properties of the connected wholes. Statistics could produce knowledge of the properties of the wholes only if it had information about statistical collectives the elements of which were wholes, i.e., if we had statistical information about the properties of many languages, many price systems, etc. But, quite apart from the practical limitations imposed on us by the limited number of instances which are known to us, there is an even more serious obstacle to the statistical study of these wholes: the fact which we have already discussed, that these wholes and their properties are not given to our observation but can only be formed or composed by us from their parts.
What we have said applies, however, by no means to all that goes by the name of statistics in the social sciences. Much that is thus described is not statistics in the strict modern sense of the term; it does not deal with mass phenomena at all, but is called statistics only in the older, wider sense of the word in which it is used for any descriptive information about the State or society. Though the term will to-day be used only where the descriptive data are of quantitative nature, this should not lead us to confuse it with the science of statistics in the narrower sense. Most of the economic statistics which we ordinarily meet, such as trade statistics, figures about price changes, and most “time series,” or statistics of the “national income,” are not data to which the technique appropriate to the investigation of mass phenomena can be applied. They are just “measurements” and frequently measurements of the type already discussed at the end of Section V above. If they refer to significant phenomena they may be very interesting as information about the conditions existing at a particular moment. But unlike statistics proper, which may indeed help us to discover important regularities in the social world (though regularities of an entirely different order from those with which the theoretical sciences of society deal), there is no reason to expect that these measurements will ever reveal anything to us which is of significance beyond the particular place and time at which they have been made. That they cannot produce generalizations does, of course, not mean that they may not be useful, even very useful; they will often provide us with the data to which our theoretical generalizations must be applied to be of any practical use. They are an instance of the historical information about a particular situation….
