1st Prize in VSP 2021: The “pretense of knowledge” as a method of governance
Economics Set Free: In Defense of Hayek
David Andrew McMillan
First Prize in Vernon Smith Prize Contest 2021.
Policymakers treat economics mysteries as a series of economics puzzles or optimization problems: If only we had more data, they claim, our predictions would be more accurate and evidence-based interventions would deliver greater benefits. The inclination to view economics as a field of scientific study must be avoided; economics is a complex field of social study. This essay advances three claims: (1) the Keynesian distinction between resolvable and radical uncertainty reveals why mainstream economists fail to predict “low-probability” economic events; (2) economic knowledge is particularized and is best communicated through the price system and markets; and (3) economics will never reveal “constants” in the same way that mathematical constants are revealed because preferences, people, and societies change. When economists and policymakers claim to possess knowledge about economic mysteries, they damage the credibility of economics; in advancing the three arguments, this essay provides a critique of bogus quantification in economic and social policy.
One dogma held by many within the scientific community is that previous scientific discoveries already explain the natural world, which leaves just the remaining
details to be worked out.1 Increasingly, economists and policymakers seem to hold the same dogma about economic and social events. While the former scientific dogma may be legitimate; the latter economic dogma is illegitimate. The relevant difference between natural sciences and social sciences is the property of non-stationarity: Economic and social relationships are subject to change over time whereas laws of nature are fixed.2
Since economic problems are not the same as scientific problems, it is unsurprising that economists have a different level of predictive capability than scientists.3 The good
news is that if we free economists from an obligation to supply governments with subjective probabilities about radical uncertainty through Bayesian inferences or
mathematical models, economics will be more trustworthy and enjoyable. The bad news is that we find ourselves in a position where we must liberate the field of
economics. The purpose of this essay is not to advance a positive argument of what economics ought to be; rather, this essay advances a negative argument about what it
ought not to be: Bogus quantification does not make sound economic policy.
The Keynesian distinction
What is the difference between risk and uncertainty? In the 1930s, John Maynard Keynes claimed that knowledge of the future is fluctuating, vague, and uncertain; when
faced with taking actions in one’s best future interest, said Keynes, it would be simply wrong to use an expected-value-adjusted cost-benefit analysis to guide behavior.4 While probability distributions in a game of roulette are clear, we cannot apply probability distributions to determine whether a particular war will occur.5 These claims from Keynes are consistent with powerful arguments that he made about probability in the 1920s.6 Keynes underscores the difference between resolvable uncertainty and radical uncertainty: Resolvable uncertainty can be resolved through research or probability distributions whereas radical uncertainty cannot be resolved by either means.7 Games of chance are well-understood: If I place a bet on a number listed on a roulette wheel, the odds of winning the payoff can be determined. However, we can only say that the probability of World War III is radically uncertain because we lack a scientific basis for arriving at a probability estimate. In response to Keynes, Frank Ramsey argued that even in the face of uncertainty, people act, and actions reflect individual calculations of events’ probability; the actions people take reflect their subjective probabilities about events.8 At first blush, it is unclear how one could measure subjective probabilities.
Ramsey claims that one could quantify subjective probabilities by using an instrument like a psychogalvanometer (e.g., a lie detector).9 By way of clarification, Ramsey holdsbthe view that when standard probability theory fails, we can and should rely on subjective probabilities to guide future action. While Keynes held that probability
theory fails to account for the probability of a future war breaking out (i.e., a radically uncertain future event), Ramsey suggests that we use lie detector tests to determine
whether a radically uncertain future event will occur. In the end, Keynes lost to Ramsey in the debate over the nature of uncertainty, so the idea of radical uncertainty
disappeared from economics for over 50 years.10 If Ramsey is correct, we should have been able to predict the financial crisis of 2007-2008. Ramsey’s view, extended to before 2007-2008, would recommend that financial institutions take a random sampling of economists, ask them to provide an estimate of the likelihood of a financial crisis, average their individual subjective probabilities, and evaluate data on past financial crises. We know that this does not work because it seems to be the approach that risk managers took in the years preceding the 2007-2008 financial crisis. David Viniar, former chief financial officer of Goldman Sachs, stated (in the early days of the financial crisis), “We were seeing things that were 25-standard deviation moves, several days in a row.”11 A 25-standard deviation event, also referred to as a 25-sigma event, is usually and incorrectly thought to be an event likely to occur every 100,000 years.12 However, the estimate of 100,000 is off by more than 130 decimal points, which led one paper to relate the likelihood of a 25-sigma event occurring with the likelihood of hell freezing over; the same researchers suggested that either financial institutions were really unlucky in 2007-2008 (in which case they shouldn’t manage other people’s money) or they are incompetent.13
A third possibility is that the 2007-2008 financial crisis was not a 25-sigma event in the real world, but it was a 25-sigma event in the model of the world constructed by
financial institutions and mainstream macroeconomists. If this is true, it would only be logical for mainstream macroeconomists and financial institutions to reconsider how
they estimate risk. In 2008, then-Chairman of the Federal Reserve Alan Greenspan asserted, “I found a flaw in the model that I perceived is the critical functioning
structure that defines how the world works.”14 Perhaps one flaw in Greenspan’s model is that stability in prices and stability in the economy are perfectly associated; similarly, another flaw might be artificially lowering interest rates—past the natural rate of interest – at the first signal of economic trouble.15 Austrian economists warned that stability in prices seemed inconsistent with stability in the economy.16 Austrian economists also warned that artificially low interest rates would lead to investments in
business projects that would not otherwise be profitable at higher interest rates.17 If thebChairman of the Federal Reserve has the humility to recognize when his view of the
world is wrong, others in government can too.
The financial crisis of 2007-2008 damaged the reputation of mainstream macroeconomists because they failed to predict the impeding financial crisis.18 Yet, macroeconomists are blamed because they supplied subjective probabilities for the occurrence of a financial crisis, an economic mystery, and something that no one should be responsible for forecasting. Blame should be conferred upon on governments for expecting macroeconomists to play the role of fortune-teller, meteorologist, and
therapist. The 2007-2008 financial crisis should have been a wakeup call: Macroeconomic research should not be in “fine-tuning” mode with the local-maximum
of the dynamic stochastic general equilibrium world, it should be in “broad exploration mode.”19 The 2007-2008 financial crisis should have been a clarion call: Consider old
and new narratives and theories about business cycles, but only accept narratives that (1) account for past financial crises and (2) avoid treating an economic mystery as an economic puzzle waiting to be optimized (as best exemplified in the actions of the Federal Reserve). The repeated failure of a model to accurately capture social and
economic behavior is not an accurate model; moreover, a sufficient justification to use a model that relies on bogus quantification is not “we need more data to be more
accurate.” Keynes is right. If we were more certain about our uncertainty, our actions would reflect how little we know about society and our own economies. We need more
experimental economics and less mathematical economics. We need to foster substantive debates about how individuals, firms, and governments behave under
uncertainty. Governments need to consider the resolvable-radical uncertainty distinction; when they do, they will realize that individuals, firms, and nations been
operating on the view that all uncertainty is resolvable when most uncertainty about social and economic conditions is not resolvable. Therefore, the government-driven
approach to collect more information about society overlooks the fact that no amount of data can ever hope to explain an economic or social phenomenon characterized by radical uncertainty.
On economic knowledge
According to Todd Zywicki, Friedrich Hayek made a distinction between two types of knowledge: Scientific knowledge and spatio-temporal knowledge.20 The
distinction is important for the purposes of understanding why governments ought not try and serve as a centralized decision maker; Hayek’s distinction also serves as a
reminder that governments cannot collect all relevant information about interactions within a market economy. Scientific knowledge, according to Hayek, is knowledge of
general and scientifically falsifiable rules.21 A panel of government economists, for example, might try to obtain and analyze all relevant economic information in order to
allocate those resources according to some rule.22 Hayek notes that (1) if we possess all relevant information, (2) we start out from a given system of preferences, and (3) we command complete knowledge of means, we could mathematically state and solve resource-allocation problems as optimization problems that follow a rule.23 However, Hayek notes, the economic problem of society is not one of scientific knowledge but one of spatio-temporal knowledge or “knowledge of the particular circumstances of time and place.”24 A panel of government economists are insufficient to coordinate the receipt of abstract and tacit knowledge from everyone within a society, integrate this knowledge, and taking actions; instead, suggests Hayek, the economic problem of society can be solved through decentralization because only then can the circumstances of time and place be used.25
Knowledge of time and place includes facts on the ground, which individuals in an economy gather as they make informed economic decisions.26 Knowledge of time
and place might include unspoken tacit knowledge that individuals might not conceptually view in clear cost-benefit terms.27 Knowledge of time and place includes
particularized information that might only make sense to particular individuals.28 We might, for example, say that the willingness to pay and the willingness to accept of
consumers and producers are guided by knowledge of time and place. When an individual decides to purchase a product or sells a product, he confers upon the
product a value whose inputs include facts on the ground, tacit knowledge, and terms that may only be particularized to him. If, instead of through voluntary market
transactions, he was required to pass this information along to the government through contingent valuation or other means, he would be unable to share all the relevant
information that guided his decision. Governments routinely fail to recognize this distinction. The epistemic pitfall that governments ignore is that markets capture more
knowledge of time and place than any other system devised; the price system allows individuals to use and rely on their knowledge of time and place to participate in
mutually beneficial exchanges. Market prices, therefore, are the best way that individuals within a society can appreciate the knowledge of others and contribute their
knowledge via the recursive process of voluntary exchange. On the one hand, Hayek’s knowledge distinction should make economists and public policymakers glad: Markets, through stated preferences and individual choices, capture knowledge of place and time. On the other hand, this seems to reveal that some government bureaucrats, whose primary responsibility is to collect economic data from individuals and firms, duplicate what markets reveal and search for information that markets already capture.
On what economics can never offer
The speed of light in a vaccum, Planck’s constant, Faraday constant, and the gravitational constant (“Big G”) are all physical constants that students might learn
about during an introductory physics course. I suppose some may wonder: Why are there no economic constants for students to learn in an introductory economics course? The short answer is that patterns in economics are “short-lived and sampledependent.”29 Preferences change, population demographics change, technologies
change, and institutions change because people change. Austrian economics is compatible with methodological individualism; collectivism simplifies individual
variation whereas methodological individualism does not.
It has been suggested that the closest scientific relative to economics is biology because the concepts of mutation and adaption as seen in biological systems are seen in
complex systems such as economies.30 Yet, since individual humans are biological organisms and since economies are comprised of individuals, any similarity between
biology and economics results because people (like other organisms) change. So, if econometricians ever came close to representing a particular type of human action
through a constant, the constant would soon change and therefore no longer be a constant; alternatively, when applied to another country through cross-sectional
variation, the constant would fail to remain “constant.” Imagine a world where physicists did not have constants, what good would variables be? Consider, the
equation E (Energy) = h (Planck’s constant) x f (frequency), which is used to determine the energy of a photon or particle. The equation has one constant (Planck’s constant) and one variable (frequency), yet without Planck’s constant the equation falls apart. Ludwig Von Mises notes that there are no economic constant relations as there are in physics; instead, economics has variables.31 Yet, since there are no constants, says Mises, it is “pointless to talk of variables.”32 The claim about economics offered by Mises is logically equivalent to the claim I advanced about physics.
Yet, an important question remains: What does the nonexistence of economic constants tell us about the field of economics? The first lesson is that economics is not a
science. This does not mean that economics should resist rationalist methods; we have a fuller understanding of markets because Vernon Smith introduced laboratory
experiments in economic analysis and millions of lives in the developing world have been improved because of Michael Kremer’s decision to introduce randomized control
trials in developmental economics. Rather, governments must understand that economists are best left to understand not predict human action. The second lesson is that
people have incredible creativity, and the economics-as-a-science view diminishes this creativity. Many governments and some economists endorse a view of human beings as mainly reactive and leaving relatively little room for creativity. If a car is approaching a man at a high speed, the man jumps out of the way: On the view of government, the man is instinctive, reactive, and the subject of prediction.33 Yet, this view of a person diminishes the entrepreneurial spirit essential to individual and social improvement: A barefoot man walks around the road, sees cowhide, and envisions making shoes.34 In this second example, human action is seen as creative and unpredictable. The absence ofbeconomic constants tells us that the view of people as instinctive, uncreative, reactive, and the subject of prediction is misguided. Third, economics will never be able to completely predict human action. If a government wanted to translate verbal intobmathematical formulas and later back into verbal economics, the complexities of thebreal world would be lost in translation.35 Government models of human action
oversimplify the complexity of individuals and, as Goodhart’s law holds, becomebirregular once a supposed regularity is established.
This section aims to hammer home the last nail in the coffin of economics as a natural or physical science. Popperian analysis draws the demarcation between nonscience and science through a standard falsification test. Science passes tests of falsification, non-science does not. Economic analysis, including neoclassical economics, does not survive the test of falsification. Most economists will concede this point.36 More importantly, “no extensive historiographical research is required to reveal that the development of economic analysis would look a dismal affair through falsificationist spectacles.”37 Therefore, it follows that economics is not a science since economics and do not pass the test of falsification. This is what Hayek meant when he said that Popper gave a test to distinguish between scientific and not scientific.38 Governments should be reminded of this fact. This essay advanced three interlocking arguments: (1) the failure of economists, governments, and firms to distinguish between risk and uncertainty causes real harm by suggesting that the future is more predictable than it is, (2) Hayek’s distinction in the types of knowledge reveals that economic knowledge is best captured by individuals interacting through markets and not through government planning, (3) the absence of economic constants reveals the field’s undervaluation of human creativity and how economics should strive to describe and not predict human behavior. Tomorrow, who
knows what governments may use data or subjective probabilities to justify? Tomorrow, who knows what governments may ask economists to justify through subjective probabilities? Today, city, county, state, and federal governments must recognize that bogus quantification does not make sound public policy, nor does it overcome the principle of non-stationarity or Popper’s falsification test. When they do, economics will be liberated in the manner Hayek intended.
David Andrew McMillan
The University of Edinburgh, Edinburgh, UK
1 Rupert Sheldrake, Science Set Free: 10 Paths to New Discovery, (New York: Random House, Inc., 2012), 6.
2 John Kay and Mervyn King, Radical Uncertainty: Decision-making Beyond the Numbers (New York: W.W.b Norton & Company, Inc., 2020), 35.
4 John Maynard Keynes, “The general theory of employment.” The Quarterly Journal of Economics 51, no. 2 (1937): 213-214. Keynes refers to what I call “expected-value-adjusted cost-benefit analysis” as
“Benthamite calculus multiplied by expected probabilities.”
5 Ibid. Keynes writes, “The sense in which I am using the term is that in which the prospect of a European war is uncertain, or the price of copper and the rate of interest twenty years hence, or the obsolescence of an invention, or the position of private wealth-owners in the social system in 1970. About these matters there is no scientific basis on which to form any calculable probability whatever. We simply do not know.”
6 John Maynard Keynes, A Treatise on Probability (London: Macmillan and Co., 1921), 31-32. Keynes states that many are wrong in saying that—in some instances (i.e., when it is cloudy)—the decision to carry an umbrella is an arbitrary decision and not one in which probability or confidence-levels apply. Keynes soon thereafter states, “Some cases, therefore, there certainly are in which no rational basis has been
discovered for numerical comparison.” While doppler radar as applied to weather has given meteorologists another tool to counsel individuals on the weather, meteorologists are often wrong. In the same way that individuals give credence to the predictive power of meteorologists, they give predictive power to economists.
7 John Kay and Mervyn King, Radical Uncertainty: Decision-making Beyond the Numbers (New York: W.W.Norton & Company, Inc., 2020), 73.
8 Frank P. Ramsey. (1926). Truth and probability. In R.B. Braithwaite, editor, The Foundations of Mathematics and other Logical essays, chapter 7, pages 156-198. McMaster University Archive for the History of Economic Thought.
10 Ibid. One can also find fault with Ramsey’s faith in subjective probability because undecidable problems are decision problems in computability theory for which an algorithm is unable to always arrive at a yes-or-no answer; Alan Turing arrived at this conclusion two years after Ramsey died.
11 Kevin Dowd, John Cotter, Chris Humphrey, and Margaret Woods. “How unlucky is 25-sigma?” The Journal of Portfolio Management 34, no. 4 (2008): 76. There is no indication that Viniar was using hyperbole: He genuinely believed that a 25-sigma event was not just entirely possible, but that it could happen several days in a row. When can we know when quantification go too far? One curt answer might be: Look to Viniar.
12 Ibid, 77.
13 Ibid, 79.
14 Committee on Oversight and Government Reform, ‘The Financial Crisis and the Role of Federal Regulators’, House of Representatives (2008), www.govinfo.gov/content/pkg/CHRG110hhrg55764/html/CHRG-110hhrg55764.htm.
15 Steve H. Hanke, “The Fed’s Modus Operandi: Panic.” Cato Institute, 18 Mar. 2009, www.cato.org/commentary/feds-modus-operandi-panic.
17 Philip Booth, “The financial crisis shows why we should admire Friedrich Hayek.” Institute of Economic Affairs, 22 Jan. 2009, www.iea.org.uk/blog/the-financial-crisis-shows-why-we-should-admire-friedrichhayek.
18 Ricardo J Caballero, “Macroeconomics After the Crisis: Time to deal with the pretense-of-knowledge syndrome.” Journal of Economic Perspectives 24, no. 4 (2010): 85-102.
20 Todd J. Zywicki, and Anthony B. Sanders. “Posner, Hayek, and the economic analysis of law.” Iowa L. Rev. 93 (2007): 559. The term ‘spatio-temporal knowledge’ is not used by Hayek, but I think it is a suitable way to describe knowledge of time and place. While I cite Zywicki, I avoid a secondary citation by noting that Zywicki made this claim about Hayek. I did verify Zywicki’s claim in Hayek’s work; however, Zywicki’s description is neat and allowed me to find Hayek’s work, so I am extending credit to him.
23 Friedrich August Hayek, “The use of knowledge in society.” American economic review 35, no. 4 (1945): 519-530.
24 Ibid, 521.
25 Ibid, 524.
26 Todd J Zywicki, and Anthony B. Sanders. “Posner, Hayek, and the economic analysis of law.” Iowa L.
Rev. 93 (2007): 559.
27 Ibid. Slavoj Žižek has referred to “unknown knowns,” which seem to be a relative of tacit knowledge. Unknown knowns are things people know to be true, but refuse to acknowledge. In some cases, it is
possible that there can be tacit knowledge that people not only just cognize, but also are afraid to verbalize for one reason or another.
29 Chakrabarti, Anindya S., and Ratul Lahkar. “Absence of economic and social constants.” The European Physical Journal Special Topics 225, no. 17 (2016): 3115-3119.
31 Ludwig Von Mises, Theory and history, (Auburn: Ludwig von Mises Institute, 1985), p. 11.
33 Israel M. Kirzner, Method, process, and Austrian economics. (Lexington, MA: Lexington Books, 1982), 15-17.
34 Ibid, 15-17.
35 Murray N. Rothbard, “Praxeology: The methodology of Austrian economics.” The Foundations of Modern
Austrian Economics (1976): 19-39.
36 Douglas W Hands, “Karl Popper and economic methodology: a new look.” Economics & Philosophy 1, no. 1 (1985): 83-99.
37 Spiro, Latsis. “Method and appraisal in economics.” Noûs 15, no. 2 (1981): 225-230.
38 F. A. Hayek, “The pretense of knowledge. Nobel Prize Speech, Salzburg, 11 December.” (1974).
Anindya S. Chakrabarti and Ratul Lahkar. “Absence of economic and social constants.” The European Physical Journal Special Topics 225, no. 17 (2016): 3115-3119.
Committee on Oversight and Government Reform, ‘The Financial Crisis and the Role of Federal Regulators’, House of Representatives (2008), www.govinfo.gov/content/pkg/CHRG-110hhrg55764/html/CHRG110hhrg55764.htm.
Douglas W. Hands, “Karl Popper and economic methodology: a new look.” Economics & Philosophy 1, no. 1 (1985): 83-99.
Frank P. Ramsey, (1926). ‘Truth and probability.’ In R.B. Braithwaite, editor, The Foundations of Mathematics and other Logical essays, chapter 7, pages 156-198. McMaster University Archive for the History of Economic Thought.
F. A. Hayek, “The pretense of knowledge. Nobel Prize Speech, Salzburg, 11 December.” (1974).
F. A. Hayek, “The use of knowledge in society.” American economic review 35, no. 4 (1945): 519-530.
Israel M. Kirzner, Method, process, and Austrian economics. (Lexington, MA: Lexington Books, 1982), 15-17.
John Kay and Mervyn King, Radical Uncertainty: Decision-making Beyond the Numbers (New York: W.W. Norton & Company, Inc., 2020), 35.
John Maynard Keynes, A Treatise on Probability (London: Macmillan and Co., 1921), 31-32.
John Maynard Keynes, “The general theory of employment.” The Quarterly Journal of Economics 51, no. 2 (1937): 213-214.
Kevin Dowd, John Cotter, Chris Humphrey, and Margaret Woods. “How unlucky is 25-sigma?” The Journal of Portfolio Management 34, no. 4 (2008): 76.
Ludwig Von Mises, Theory and history, (Auburn: Ludwig von Mises Institute, 1985),p.11.
Murray N. Rothbard, “Praxeology: The methodology of Austrian economics.” The Foundations of Modern Austrian Economics (1976): 19-39.
Philip Booth, “The financial crisis shows why we should admire Friedrich Hayek. ”Institute of Economic Affairs, 22 Jan. 2009, www.iea.org.uk/blog/the-financialcrisis-shows-why-we-should-admire-friedrich-hayek.
Ricardo J Caballero, “Macroeconomics After the Crisis: Time to deal with the pretense-of-knowledge syndrome.” Journal of Economic Perspectives 24, no. 4 (2010): 85-102.
Rupert Sheldrake, Science Set Free: 10 Paths to New Discovery, (New York: Random House, Inc., 2012), 6.
Spiro, Latsis. “Method and appraisal in economics.” Noûs 15, no. 2 (1981): 225-230.
Steve H. Hanke, “The Fed’s Modus Operandi: Panic.” Cato Institute, 18 Mar. 2009, www.cato.org/commentary/feds-modus-operandi-panic.
Todd J Zywicki, and Anthony B. Sanders. “Posner, Hayek, and the economic analysis of law.” Iowa L. Rev. 93 (2007): 559.