Economists rely on Vector Autoregression (VAR) models to forecast macroeconomic time series that may infer the effects of structural shocks and estimate unobservable cyclical components of macroeconomic aggregates. A VAR model is made up of a system of equations that represents the relationships between multiple variables. For example, variables such as unemployment, interest, and inflation rates may be incorporated into a VAR model. Theories have been developed that claim to improve model misspecification, thereby avoiding inaccuracies.1Francesca, Loria, et al. “Economic theories and macroeconomic reality”. Journal of Monetary Economics, vol. 126, 2022, pp. 105-117, https://doi.org/10.1016/j.jmoneco.2021.12.001. These models are interesting but the substrate on which any economic framework rests should take priority. This analysis argues that in the absence of classical economic theory, a model, regardless of complexity, cannot correspond to economic realities or provide real-world solutions. The Keynesian econometric framework, which nearly all models are based on, possesses fallacious twentieth-century concepts regarding centralized planning and a demand-driven economy.
The well-known regression techniques associated with the use of a Gibbs sampler or Bayesian statistics integrated and exploited in new VAR models are not in question. These statistical models are useful but the underlying assumptions (ex-ante) should be based on classical economic theory before the theoretical structure (ex-post) can be expected to yield real solutions. As Jean-Baptiste Say stated, “The study of statistics may gratify curiosity, but it can never be productive of advantage when it does not indicate the origin and consequences of the facts it has collected; and by indicating their origin and consequences, it at once becomes the science of political economy.”2Say, Jean-Baptiste. A Treatise on Political Economy. 1836. Cosimo Classics, 2007. Economic theory cannot be based solely on statistical data because human affairs and their economies are too complex. Statistical models cannot provide cause and effect, without incorporating what Say referred to as political economy. He articulates how Adam Smith used a combination of statistics and political economy to develop his theory. Although, the term statistics was not defined precisely as it is now. Statistics in Say’s time referenced the numerical data collected and did not include probability as our contemporary models do. Today, talented statisticians can manipulate data to satisfy the politically influenced economic system.
It’s important to notice how solutions are defined specifically when integrating statistical models with the Keynesian apparatus. Solutions are defined by a workable model, not by actually solving the problem within the physical framework. Models containing GDP variables further exasperate forecast reliability. This is especially true when projections are made based on GDP gaps. Determining potential GDP is controversial as there is a lack of consensus among economists as to how it should be measured. Generally, potential output is based on full employment. The forecast can be enigmatic, e.g., in 2020, during the height of COVID lockdowns, the Federal Reserve Bank of St. Louis had projected the fourth-quarter potential GDP to be $19.41 trillion. However, the actual GDP was $21.48 trillion.3Majaski, Christina. “Output Gap: What It Means, Pros & Cons of Using It, Example.” Investopedia, 30 Aug. 2021, https://www.investopedia.com/terms/o/outputgap.asp. The positive GDP gap is puzzling considering many sectors of the U.S. economy had cut or ceased production and many employees were at home.
Keynesian economic theory is the primary influence on any model, and the VAR is no exception. The VAR model forecast is based on prior results (e.g., VARIABLEt-1). The formulas connected with the data may be producing results satisfactory to a statistician’s expectations. The question is how useful are these workable models on the macro side? Some of the latest research suggests significant improvements have been made in forecasting by noting the marginal possibility of conjugate priors commonly used and embedding the VAR into a Gibbs sampler to allow for multiple measurements from each economic variable.4Francesca, Loria, et al. “Economic theories and macroeconomic reality”. Journal of Monetary Economics, vol. 126, 2022, pp. 105-117, https://doi.org/10.1016/j.jmoneco.2021.12.001.
The microeconomic side or managerial side in manufacturing may benefit more from VAR models, but lean production methods have ensured surpluses will be essentially nonexistent. The nonexistent surplus has created a plethora of issues beginning with the lockdowns in 2020. These critiques parallel with the illustration that Steven Kates articulated revealing the false dichotomy between a downturn in the economy and demand.5Kates, Steven. “Why Keynesian Concepts Cannot Be Used to Explain Pre-Keynesian Economic Thought: A Reader’s Guide to Classical Theory.” Quarterly Journal of Austrian Economics, vol. 17, no. 3, 2014, pp. 313-326. ProQuest. The equations, data, and model may be sound and workable, but the false assumptions based on faulty economic theory often fail to return real-world solutions.
All modern economists who specialize in the study of macroeconomics have been inundated with Keynesian thought. Keynesian models are structured around the belief that variations in aggregate demand are the basis for explaining variations in output and employment. Aggregate demand is distinct from effective demand (a term that dates back to the eighteenth century) which is concerned with what must occur when an individual decides to purchase products. Aggregate demand was brought into economic discourse in 1936 by John Maynard Keynes. This definition included all individual demands within an economy.6Kates, Steven. “Why Keynesian Concepts Cannot Be Used to Explain Pre-Keynesian Economic Thought: A Reader’s Guide to Classical Theory.” Quarterly Journal of Austrian Economics, vol. 17, no. 3, 2014, pp. 313-326. ProQuest.
Keynes focused on monetary aggregates, i.e., saving and investing were thought of in money terms. Value is not considered, or what some classical economists refer to as the real side of the economy—production. Economic growth in terms of production is replaced by this new focus on the monetary. National saving was compared with individual saving. Savings on an individual level were viewed as detrimental to the economy, and those who spend all their income were viewed as providing stimulus to the economy. Therefore, spending was preferred over savers who hoarded income, creating an economic recession. Keynes’ theory holds that centralized planning by governments could control a nation’s spending, thereby stimulating the economy via spending. The Keynesian theory insists that economic downturns occur because of unplanned hoarding by individuals or banks. The shrinkage in aggregate demand will influence production, i.e., it will encourage reductions in aggregate supply. Only government spending can create the necessary conditions for returning to equilibrium between aggregate demand and supply.
The General Theory from Keynes is prevalent in contemporary macroeconomic practice, e.g., the current trend combines government spending with monetary policy via the Federal Reserve attempting to lower wages and increase unemployment. This combination is believed to address two problems, recession and inflation. The spending will return demand levels and pull the nation from a possible recession and the Fed’s policy will provide an avenue for wage reduction and higher unemployment which will reduce inflation. Many have noticed that reducing wages and raising unemployment seems antithetical to creating conditions for a better economy.7Schiff, Peter. “Bond Bear Will Maul Stocks and the Dollar.” Zerohedge, (July 2023), https://www.zerohedge.com/markets/peter-schiff-bond-bear-will-maul-stocks-and-dollar.
Keynes’ theory has had serious problems from its inception. It is fallacious to think of national savings in such terms (monetary). Nations do not save by hoarding money in financial institutions as individuals may. A nation’s savings consists of raw material and productive assets in existence that can be used to transform inputs into a final output for future returns. The nation’s potential to produce a product constitutes its savings, it has little to do with the monetary. The nation’s production capacity is the real side of the economy understood by classical economists.
Jean-Baptiste Say, near the turn of the nineteenth century (c. 1803), composed a treatise that centered on a concept i.e., all economic activity began on the supply side. This became known as Say’s Law of Markets and it was universally accepted until 1936.8Kates, Steven. “Why Keynesian Concepts Cannot Be Used to Explain Pre-Keynesian Economic Thought: A Reader’s Guide to Classical Theory.” Quarterly Journal of Austrian Economics, vol. 17, no. 3, 2014, pp. 313-326. ProQuest. It was not part of the new science of economics taught by classical economists of the time, but, as Ludwig von Mises elaborated it was “a preliminary—the exposure and removal of garbled and untenable ideas which dimmed people’s minds and were a serious obstacle to a reasonable analysis of conditions.”9Mises, Ludwig von. “Lord Keynes and Say’s Law [of Markets: The Contribution of Each to the Theory of the Trade Cycle].” Freeman (Irvington-on-Hudson N.Y.), vol. 1, 1950, pp. 83– 85. Prior to Adam Smith and Say’s theories, there were two reasons given for a business failure: the scarcity of money and overproduction. In a passage from his Wealth of Nations, Smith exposed the fallacy regarding money scarcity. For this section, however, we are concerned with Say’s treatise and what proved to be a myth regarding overproduction and business failure.
There can never be an absolute overproduction of economic goods. There will always be unsatisfied needs that a larger supply of certain economic goods can satisfy. There may, of course, be relative overproduction. This stems from an error in management where a certain good underperforms compared to its related counterpart. The management has misread market conditions and produced too much of one good and likely too little of another. In Mises’s analysis of Say’s Law, he provided the example of shirts and shoes. If consumers are purchasing a larger quantity of shoes and a smaller quantity of shirts, this is not an indication of overproduction of all commodities.10Mises, Ludwig von. “Lord Keynes and Say’s Law [of Markets: The Contribution of Each to the Theory of the Trade Cycle].” Freeman (Irvington-on-Hudson N.Y.), vol. 1, 1950, pp. 83– 85. It is rather an error in management, they have failed to understand the needs or desires of their customers, i.e., they have failed to anticipate the conditions of the market. A classical economist would never interpret such a scenario as a general depression nor would they consider it applicable to aggregate demand. The situation is improved by reversing the ratio of production between commodities. Here is where an argument can be made for VAR models. In manufacturing scenarios, where forecasts are based on priors, these models reveal trends that can assist managers with understanding future market conditions for correct production ratios. Again, the goal here is not to discourage the use of such models on a micro-scale. Problems emerge when models are used in conjunction with Keynesian theory, or when they are used to justify overall production cuts.
Say and Smith dispelled the myth that periods of bad business were caused by scarcity of money and absolute overproduction. Many classical economists have demonstrated that reoccurring depressions of trade were caused by repeated attempts to stimulate the economy via government spending and credit expansion. They did, however, concede that credit expansion would create an initial boom in business, but it would inevitably be followed by a bust. Their theories held until the so-called Keynesian revolution took precedence (some economists argue it wasn’t really a revolution). True statesmen could see value in classical economic theory for the prosperity of their nation but the career politician found no use for it. Or, as Mises eloquently stated, “It could not influence demagogues who care for nothing but success in the impending election campaign and are not in the least troubled about what will happen the day after tomorrow.”11Mises, Ludwig von. “Lord Keynes and Say’s Law [of Markets: The Contribution of Each to the Theory of the Trade Cycle].” Freeman (Irvington-on-Hudson N.Y.), vol. 1, 1950, pp. 83– 85. These politicians build their wealth from wars at the expense of the proletariat and persevere their power via centralized planning. Free markets have no place in their artificial economy constructed of fiat money and empty promises of prosperity through spending.
Moreover, the aggregate demand fallacy and its solution, i.e., government spending, were policies already adopted by politicians before Keynes’s book, General Theory, was published. According to some economists, Keynes wrote it as an apologia for the policies already in place by governments.12Mises, Ludwig von. “Lord Keynes and Say’s Law [of Markets: The Contribution of Each to the Theory of the Trade Cycle].” Freeman (Irvington-on-Hudson N.Y.), vol. 1, 1950, pp. 83– 85. Keynes’s work provided the semblance of a scientific allure that could be touted before their constituents. It also provided credence for the mathematical economist who treasures the workable model over a reality-based solution.
In consideration of the classical economic view, it may be beneficial to note the economic calamity that has begun to unfold since 2020. The Keynesian theory is well integrated into the universities which have produced managers who acquire positions at major manufacturing facilities in the United States. Upper management seems to prefer mid-managers who boast of their credentials, proudly displaying their black belt certification in Six Sigma. The JIT lean production methodology is compatible with the aggregate demand theory. This philosophy emerged from Japan after World War II and found its way to the U.S. in the 1970s. The focus is not simply looking for ways to improve cost by creating more efficient production systems. A portion of the process entails cutting storage costs by eliminating surplus. Auto manufacturing adheres to these JIT practices as they do not store large surpluses of parts. The parts they need to complete the production process arrive just in time and are immediately incorporated into the assembly.
VAR models can certainly improve forecasts but they should be understood through the lens of relative supply and effective demand. This is especially true for manufacturing when it comes to production ratios. During the 2020 lockdowns, products were redirected to areas that could sustain relative supply and effective demand. The market conditions were changing, artificially via government intrusion, but changing nonetheless. Auto manufacturing, for a time, ceased and semiconductors that were being produced for new vehicles were repurposed and made readily available for electronic devices. Their ability to improvise is noteworthy, but they were short-sighted regarding another sector, they should have continued producing for auto manufacturers as well. This would have resulted in relative overproduction but it would soon be alleviated once the government took its hand off the scale.
If nothing else becomes of the 2020 totalitarian exercise, it should at least prove the nation’s production apparatus is the real side of the economy, i.e., supply drives economies, and aggregate demand is a fallacy. Moreover, government spending has caused significant inflation and its interference in production has crippled the supply chain. Although had there been a substantial surplus many of the businesses may have weathered the crisis far better. The theoretical structure that Jean-Baptiste Say and Adam Smith helped develop could have spared humanity from the disastrous failed Keynesian-based economy, but they are viewed by the ruling class as antiquated and antithetical to government. In truth, governments will not likely retire Keynesian economic theory, as too much is at risk for the politician. Relinquishing control of the economy where government officials have the power to essentially choose winners and losers via illusionary mechanisms of a failed theory, as opposed to allowing a truly free market to persist, is incomprehensible for the politician.