Facts and Economic Science

It is clear that economics is in chaos when there are significant disagreements about basic facts of economics. There is an abundance of statistical data presented by the media, but there are few agreed facts. Is there too much government deficit spending or too little? Is there too much debt or not enough? Is the danger inflation or deflation? The lack of agreed facts about inflation has led to divergent monetary policies of central banks, as a senior OECD economist noted recently (White, 2014):

The Bundesbank, for example, is fighting the threat of high inflation, whereas the Fed is more concerned about the prospect of deflation.

White (2013) summed up the situation earlier:

there is, in practice, no body of scientific knowledge (evidence based beliefs) solid enough to have ensured agreement among central banks on the best way to conduct monetary policy.

Economics is not science because much of economic knowledge is not based on facts derived from the scientific method where facts make or break theories. Politicians, central bankers and economists create policies based on economic theories which they have learned at university. The economic theories in textbooks (e.g. Marshall, 1890; Mankiw, 2009; Varoufakis, 1998; Hill and Myatt, 2010; Keen, 2011) are based on narratives, not facts.

Indeed university education actually teaches policy makers to ignore facts – leading to a “trained incapacity” to recognize facts. For example, the observed fact – that an economy can collapse or develop periodic crises – is not considered as fact in most textbooks which assume collapse and crisis are negligible or unimportant or unknowable (Sy, 2012). This may explain why there have been few new economic ideas or policy changes several years after the global financial crisis (GFC).

More widespread and extreme pursuit of policies based on unscientific theories has been putting the global economy in unprecedented risk, because the results have been counter-productive. Both the economic profession and economic education need urgent and radical reform, without which our economic future will be bleak. To understand how economics has been trapped in a practice of unscientific, “observation-less theorizing” (Bergmann, 2009), we indicate the origin of the practice and compare it to the scientific practice of physics.

Economic Scholasticism

Economics is not science because its practice originated in the rationalist-scholastic tradition going all the way back at least to Aristotle who created many economic terms. The rationalist-scholastic tradition believes that reason, innate knowledge and the use of dialectical methods are the chief sources of new knowledge. This tradition is still practised today by a majority of economists, whose point of departure for their research is referencing the philosophies or theories of other economists within a recognized school (e.g. Neoclassical, Keynesian, Marxist, and Austrian). The rationalist-scholastic approach leads to an epistemology based on rhetoric, arguments and disputes, which are typically unresolved in a broader economic context, leading to a pluralism of economic schools – a pluralism of dogmas.

For the school-men, only books and research papers published within a particular school are regarded as knowledge by that school. For example, it is not permissible to introduce the concept of irrationality in neoclassical economics, which believes rationality in individual decision-making is axiomatic in how an economy works.   Because of the way economic journals are segregated into different schools, criticisms of a particular school are never published in that school’s journals, where they could be most useful in stimulating progress for that school. This insularity leads to a view of economics as religion (Nelson, 2001).

Doing economics still largely means “sitting in a chair and making it up” (Bergmann, 2009) based on different paradigms, even if the theories are made-up with mathematics. Those economists who have attempted to break away from Lakatos’ “degenerative research program” of rationalist-scholasticism have not done science, but have merely imitated science, as they have been misguided by certain philosophers of science.

Imitation of Science

Economics has always envied physics and has sought to imitate it (Mirowski, 1989) and, in recent decades, through mathematization and data analysis. The neoclassical school through its concept of equilibrium has lent itself to the mathematization of Walras and Arrow-Debreu (1954).   This has helped neoclassical economics to distinguish itself from other schools in apparently gaining the status of science. But neoclassical economics is not science.

The mathematization of neoclassical economics does not change the scholastic nature of economics, because mathematics is merely used to deduce logical conclusions through mathematical arguments – another form of rhetoric. Neoclassical economics is not science if its axioms or the conclusions of its theory do not correspond to economic facts of observation. For this reason, even today, most mathematical models or theories in economics or finance are not science, because their relations to facts are often contradictory or tenuous at best – the dynamic stochastic general equilibrium (DSGE) models being important examples which are misguiding central bank policies.

The failure of mathematization to produce anything useful in economics has been blamed on mathematics itself and formalism (Blaug, 1998; Lawson, 2012). But it is the general mathematical incompetence of economists which has allowed mathematics to overwhelm the subject, as Debreu (1991) noted:

The spread of mathematized economic theory was helped even by its esoteric character. Since its messages cannot be deciphered by economists who do not have the proper key, their evaluation is entrusted to those who have access to the code. But acceptance of their technical expertise also implies acceptance of their values.

Most economists do not have the key to decipher mathematical theories and are therefore incapable of critically evaluating their contents and rejecting them. This incompetence has led to economists and the public confusing mathematical theory with scientific theory.

Unscientific mathematical models and theories have been allowed to spread “helped even by its esoteric character”, as Leontief (1970) noted:

Continued preoccupation with imaginary, hypothetical, rather than with observable reality has gradually led to a distortion of the informal valuation scale used in our academic community to assess and to rank the scientific performance of its members. Empirical analysis, according to this scale, gets a lower rating than formal mathematical reasoning.

It should be emphasized that “preoccupation with imaginary, hypothetical, rather than with observable reality” is not just a plague on mathematical theories, but virtually all economic theories, due to the rationalist-scholastic tradition mentioned above. But mathematical theories have an aura of science and therefore are less easily rejected by those less mathematically competent.

The mathematical incompetence extends to econometrics – which is the statistical analysis of economic data. Econometrics has not produced significant facts for economics, as Summers (1991) summarized its failure in macroeconomics as creating a “scientific illusion in empirical macroeconomics”. Decades of econometric studies have not discovered significant facts which economic theory must explain (not even the Phillips curve relationship). Applied econometrics typically leads to results which are neither statistically significant nor economically significant. A partial explanation may be that economists do not really understand mathematical statistics (Leamer, 1983; Swann, 2009, 2012; Ziliak and McCloskey, 2004).

Economics is not science because it has put the theoretical cart before the empirical horse.

Facts in Physics

Science is based on empiricism, not rationalist-scholasticism. Empiricism is primarily about facts with theory playing a secondary role in organizing facts and in the discovery of new facts. Facts can falsify theories, but theories cannot falsify facts – only new facts can falsify old facts. Theory begins with old facts and ends with potentially new facts. To be sure, identifying facts is not always straightforward, as there are different types of facts: simple facts, specific facts, compound facts, general facts, etc.  Scientists spend most their efforts on discovering and validating facts.

The theories of Newton and Einstein are not rationalist achievements as commonly believed judging from their deductive-axiomatic presentations – rather, they have their essential origins in empiricism. It was observed facts and their reconciliation which led them to their theories. They did not simply sit in their armchairs and dream up their theories from innate knowledge in the rationalist-scholastic tradition. Newton performed many physics experiments and was well-known as an alchemist, while Einstein invented devices with 19 patents and was involved in gyromagnetic experiments even while he was working on the theory of relativity (Galison, 1982).

The theoretical genius of Newton and Einstein consists of understanding which facts are most important among a plethora of other facts. They were also able to create new theoretical concepts to explain how those key facts are logically linked. Most people only focus on the mathematical nature of their works without understanding the importance of their empirical foundations. In a pithy repudiation of rationalist-scholasticism, Einstein (1934) said:

Pure logical thinking can give us no knowledge whatsoever of the world of experience; all knowledge about reality begins with experience and terminates in it.

 Newton’s Giants

In a letter to Robert Hooke in 1676, Newton himself provided the clue to his genius by saying famously

If I have seen further it is by standing on the shoulders of giants.

In addition to agreeing with the common interpretation that science advances by building on what went before, it is emphasized here that Newton’s theory was based on the empiricism of giants, particularly Galileo and Kepler, who in turn had built on the empirical observations of other giants such as Tycho Brahe, Copernicus, and even Ptolemy. All told, Newton’s theory rested on centuries of experimental and astronomical observations.   It was not the apocryphal apple which fell on Newton’s head that set off the “light bulb” moment of a rationalist epiphany.

Newton’s genius was to recognize the key empirical facts which needed to be linked and to be explained with a new theoretical concept. Newton’s law of gravitation states that two bodies attract with a force proportional to their masses and inversely with their distance squared. This law forms the axiomatic foundation for explaining and deducing Kepler’s empirical laws.

It would be wrong to think Newton discovered “rationally” the law of gravitation first and then deduced from it Kepler’s laws and elliptical orbits later.  It was just the reverse. Kepler’s laws and Galileo’s constant acceleration of falling bodies were used to deduce the law of gravitation. The deductive, axiomatic presentation of the Principia (Hawking, 2002, pp. 725-1160) could mislead the casual reader into thinking that it was rationalist achievement, like Euclid’s theorems. Nothing can be further from the truth.

Einstein’s Giants

Einstein also stood on the shoulders of empirical giants: Maxwell and Michelson-Morley, whose observations formed the foundation of the special theory of relativity.

James Clerk Maxwell’s genius was to summarize centuries of empirical discoveries about electricity, magnetism and their interactions as electro-magnetism in just four partial differential equations, describing electrostatics (Gauss’ law), static magnetism (Gauss’ law for magnetism) , dynamic electricity (Ampere’s law) and dynamic magnetism (Faraday-Maxwell’s law of induction).   There were large numbers of experimenters who contributed to the facts of observation behind Maxwell’s equations, including Gilbert, Franklin, Galvani, Volta, Oersted, Ohm, Ampere, Faraday, Hertz, etc., even without mentioning the ancients.

Under rationalist doctrines of Aristotle and Descartes, even in the early twentieth century most believed light must travel through a medium (like sound through the air), which was called the plenum, an invisible substance permeating the universe. If there were such a medium through which light propagates, then due to motion of the earth through the medium, the speed of light must vary depending on direction, whether it is with, against, at right angles, or at other angles relative to the motion of the earth.

In experiments performed in 1881 by Michelson, in 1887 by Michelson and Morley, and between 1902-1904 by Morley and Miller, the measured speed of light was constant to within experimental error, and detection of the medium (then called luminiferous aether) failed. The significance of the null result was appreciated by Einstein, who saw that the aether idea was unnecessary to the explanation of the facts.

Einstein postulated the principle of relativity which assumes (Hawking, 2002, pp.1167-1190) that Maxwell’s equations and the speed of light are formally the same for all frames of reference independent of their relative uniform motion. It should be emphasized here that the theory of relativity was not the result of brilliant mathematical deduction in the rationalist tradition – it was not the product of “pure reason” made up sitting in a chair – as Einstein was driven by experimental facts in an empiricism which needed to be reconciled by a new theory.

Natural Sciences

Facts are so important in physics that nearly 90 percent of publications in a survey (Morgan, 1988) are on empirical analysis and the rest are on theory and modelling. The special theory of relativity and its axioms have not altered in any significant way for over one hundred years, but the facts of the theory have been constantly enlarged, tested and refined. For example, with ever increasing technological sophistication over the years, many different experiments have been conducted to test and re-test the constancy of the speed of light, which was last measured to be accurate to less than one part in one hundred quadrillion in 2009.

As science moves to more and more complex systems, where the elementary units are more complex – e.g. molecules in chemistry, DNA in genetics, cells in biochemistry, organisms in biology, organs in anatomy, humans in medicine, etc. – more and more work consists of discovering and organizing new facts, rather than deductive theorizing. For example, nearly 100 percent of publications in chemistry are on empirical analysis (Morgan, 1988).

Perhaps due to the nature of complexity, even small changes or dynamics can lead to many different forms, possibilities and outcomes. For example, except for very simple cases, there is no deductive theory able to predict the products of a chemical reaction of two different large molecules, even though there are some physical laws such as those of thermodynamics. It is not clear whether there is, or can be, any biological laws (Dhar and Giuliani, 2010; Dorato, 2011; el-Showk, 2014), without which non-trivial theoretical deductions about biological phenomena are virtually impossible.   Natural sciences, outside physics, may have universal organizing principles such as natural selection in evolutionary biology, but may not have significant universal laws from which to indulge in deductive theorizing.

One of the goals of science may be to discover universal laws to create deductive theories, but to say science must have universal laws is a common philosophical fallacy, e.g. as Popper (1935, p.37) stated erroneously:

The empirical sciences are systems of theories. The logic of scientific knowledge can therefore be described as a theory of theories. Scientific theories are universal statements.

If this were the definition of science, then only physics is science, not chemistry or biology and certainly not medicine – an absurd conclusion. On the contrary, the empirical sciences are systems of facts. Philosophy and philosophy of science are not where economists should learn about science. Most natural sciences, outside physics, do not have significant deductive theories, but are mainly concerned with discovering and organizing facts of observation. The main activity of scientists is establishing and confirming facts and not deductive theorizing.

Economic Theories

Economics consists mainly of economic theories which are regularly falsified by facts. Economists have confused deductive theorizing, particularly using mathematics, as science. Under the rationalist-scholastic tradition, economists have not bothered much with facts, but have been theorizing based on deductions from “innate knowledge” expressed as economic laws: the law of demand, the law of supply, the law of diminishing marginal utility, etc. But these laws are not laws of science based on well-established empirical facts.

Economic laws are actually assumptions or working hypotheses which are often factually false. For example, as a basic law of economics, the law of demand is regularly violated in financial markets. Many of these markets do not have downward sloping demand curves. As the price of a stock rises its demand often increases rather than decreases (and vice versa), contradicting the law of demand, in an empirically well-established “anomaly” called the momentum effect in modern finance (see, e.g., Schwert, 2003). Yet a downward sloping demand curve is still taught in economic textbooks as though it is a universal law (with Veblen goods and Giffen goods as exceptions).

Economic laws are not established facts and hence deductions from them, even with sophisticated mathematics, have dubious validity. For example, the DSGE models used by central banks have very poor forecasting records. The credit risk models used by rating agencies, too-big-to-fail banks and regulators are based on false assumptions (Sy, 2008) and the global financial system is based on the false assumption that risk management is a science where the amount of risk capital is assumed to be accurately calculable. It is likely that all these fallacies gave rise to policies which caused the GFC.

The global financial crisis has not been properly understood as a falsification of the basic premises of economic and finance theories. The quadrillion over-the-counter (OTC) derivatives market has been built on the same false premises and has been allowed to grow unregulated. These developments clearly demonstrate that economists ignore facts, even those which are as dramatic as economic and financial collapses.   Economists have made little effort to research those facts or to analyze their significance for their economic theories – there has been nothing radically new theoretically in the several years since the GFC.

Conclusion

A revolution is required to make economics a science. Economists have been misguided by philosophers of science into pursuing a false agenda where “observation-less theorizing” in the rationalist-scholastic tradition has been its primary activity (Bergmann, 2009). Mathematical theory has been confused with scientific theory. The lack of attention to the discovery of significant economic facts has led to most effort being directed to empty theorizing which has not been useful for economics.

To make economics a science it is necessary to follow the successful examples of science, where even the greatest achievements in theoretical physics have been based on substantial empirical foundations (Einstein, 1934). An economic empiricism needs to be established with the empirical horse, in front, pulling the theoretical cart along behind it and not the other way around - it is wrong to put the theoretical cart before the empirical horse. A set of key economic facts needs to be discovered which challenges theory to explain in the context of a new scientific economic paradigm.

Reform of the economic profession and education requires a revolutionary shift from the rationalist-scholastic tradition based on made-up theories to a scientific tradition based on observed facts. The reason that economics is not science has been reduced to a single cause: lack of facts, because facts have so far been neglected relative to theories. The deficiency of facts can be overcome now more easily with advancing information technology and with proper education, provided the cause of the malaise is recognized.

More details on reforms and a full version of this post are contained in a research paper.

References and citations

This entry was posted in Methodology, Science. Bookmark the permalink.

3 Responses to Facts and Economic Science

  1. gmps2011 says:

    Well said!

    Your remarks remind me of two observations by great thinkers that have shaped my views about what is wrong with economics:

    (1) The beautifully succinct observation by Peter Wiles (Economics in Disarray) that the main thing wrong with economics is its disrespect for fact.

    (2) Norbert Wiener’s observation (God and Golem Inc.) that you cannot have a scientist like Isaac Newton until you first have a scientist like Tycho Brahe – renown for his painstaking empirical observation of facts. Wiener also said that he had seen no economists remotely comparable to Tycho Brahe.

    • admin says:

      Thanks Peter,

      Norbert Wiener has interesting comments on science, economics and econometrics in Chapter VII of God and Golem, Inc. For example:

      The success of mathematical physics led the social scientist to be jealous of its power without quite understanding the intellectual attitudes that had contributed to this power.

  2. admin says:

    Recently, I found a paper titled What economists do published in 2011 in the Journal of Applied Economics by the 1995 Nobel Laureate, Robert Lucas, who is a leader and a staunch proponent of neoclassical economics.

    It is interesting that this paper was the commencement address delivered at the University of Chicago on December 9, 1988. The fact that it was an invited paper published 23 years later in the journal indicates that they felt a need to reaffirm the Chicago method.

    The rationalist-scholastic approach of the Chicago school is made evident when Lucas (2011) said:

    ...I’m not sure whether you will take this as a confession or a boast, but we are basically story-tellers, creators of make-believe economic systems.

    At least Lucas tells the truth which escapes most other economists who do not even realize or admit that they are also telling stories. His explanation for adopting the approach is given in the conclusion of the paper.

    In any case, that is what economists do. We are storytellers, operating much of the time in worlds of make believe. We do not find that the realm of imagination and ideas is an alternative to, or a retreat from, practical reality. On the contrary, it is the only way we have found to think seriously about reality.

    In a way, there is nothing more to this method than maintaining the conviction (which I know you have after four years at Chicago) that imagination and ideas matter. I hope you can do this in the years that follow. It is fun and interesting and, really, there is no practical alternative.

    Neoclassical story-telling is done with mathematical models coupled with calibration and empirical estimation of model parameters. But economic data are not used to test their assumptions or conclusions. Not knowing the difference between these two uses of data, most economists have difficulty in denying or refuting the scientific status of neoclassical economics. We are continuing to suffer the consequences of this ignorance.

Leave a Reply