The limits of empirical economics
According to positive empiricism, adherence to economic facts is the only way to validate economic theories.
“Viewed as a body of substantive hypotheses, theory is to be judged by its predictive power for the class of phenomena which it is intended to “explain.” Only factual evidence can show whether it is “right” or “wrong” or, better, tentatively “accepted” as valid or “rejected.” …. the only relevant test of the validity of a hypothesis is comparison of its predictions with experience.”
Economic theory is to be judged by its predictive power supported by empirical regularity. Only factual or empirical evidence can make a theory go from a “meaningless” hypotheses to part of man’s accumulated knowledge. The only relevant test of the validity of a hypothesis is its predictive powers.
Friedman expanded on Popper’s “falsification principle” that a theory can never be verified but can be falsified if one finds one counter-instance that the theory fails.
“The hypothesis is rejected if its predictions are contradicted (“frequently” or more often than predictions from an alternative hypothesis); it is accepted if its predictions are not contradicted; great confidence is attached to it if it has survived many opportunities for contradiction. Factual evidence can never “prove” a hypothesis; it can only fail to disprove it, which is what we generally mean when we say, somewhat inexactly, that the hypothesis has been “confirmed” by experience” 2
Economist have taken positive empiricism one step beyond and formulated economic theory from empirical regularities; Okun’s law, the Phillips curve or, more recently, the popularity of Rogoff and Reinhart’s debt-to-GDP tipping point. Economists, also, have little restraint in throwing up graphs showing empirically causal relationships between economic variables.
This is, however, taking economics way beyond what it can do, yet, few professional economists take to the airwaves to denounce this bastardisation of the science. Empiricism can support an economic theory, but it cannot prove or disprove an economic theory.
Over 100 years ago, the limits of empiricism in economics were made crystal clear. In the article, “the elasticity of the demand for wheat”, R.A Lehfeldt (1914) attempted to determine the elasticity of demand by examining historical data of the price of wheat against the consumption of wheat. He attempted to correct for changes in other factors (ceteris paribus) and he found the elasticity of the demand for wheat to be a positive 0.6. Should we conclude from this study that the demand curve for wheat is, in reality, upward sloping? Hasn’t this empirical study shown that economic theory is wrong? Any sensible economist would explain that what is observed are not points on a stable demand curve, but ever changing intersection points between demand and supply or points moving toward such equilibria.
A demand curve is like a photograph: It is only valid for that instance since other factors change constantly so that the position of the curves is different from one instance to the next. It is impossible to empirically measure the slope of a demand curve. Although the author tried to correct for shifts in the demand and supply curves, that are too many factors (some unmeasurable) changing for him to empirically control for all of them. His task was impossible and economists should have drawn important conclusions from this failure.
Empirical estimates during this time period would also have found an overwhelming number of positive sloped demand curves. This was a period of increasing populations and monetary growth (both measured inaccurately). Hence, following positive empiricism, are we to conclude that demand curves are “rejected” as downward sloping and “confirmed” as upward sloping?
This brings up another important point that is somewhat missing from the debate over Friedman’s 1953 paper: the measurement problem in economics. Take, for example, the relationship between monetary growth and inflation (a relationship failing under the “falsification principle”). What is money and what is inflation? We have multiple definitions of money from M0, M1, M2, M3, M4 etc. Inflation is usually calculated with either the CPI or GDP deflator. Yet we know by construction that the CPI is inaccurate. We must use base period weights to focus on prices. This automatically overweighs rising prices and underweights falling prices.
Also, the original quantity theory of money related money to the prices of all transactions: anything money can buy, food, stocks, bonds, real estate jewellery. An index of this correct measure of price inflation is impossible to calculate since the weights are not calculable. Yet, the CPI or GDP deflator are inadequate proxies for this correct measure of inflation. We see today that a tame CPI is blinding central bankers to the distortive effects of asset price inflation.
In this light, positive empiricism in economics is very limited and in many cases useless.
So what is the economist to do? He goes back to theory, realising that empiricism is there to assist theoretical work but not to be confused with the foundation or replacement of economic theory.
Economics is a social science built on irrefutable axioms of human actions. Empiricism in economics is much more limited than in the physical sciences. Its only role should be to support theory.
Although John Stuart Mill was an empiricist, he was right when he said that the whole of economic science is “hypothetical.” It is a science of tendencies only.
 Milton Friedman, “The Methodology of Positive Economics”
 David Brat, “Milton Friedman’s Positivism and the Method of Economics”