To the editors:
In his essay, Henri Lepage notes that recovery from the 2007 recession was significantly more sluggish than recoveries in earlier times, including for the Great Depression of the 1930s. Lepage is correct in making this observation, as well as in criticizing the “do the same thing again, only do more of it” remedies, such as quantitative easing, that prominent theorists such as Lawrence Summers have advanced.
Modern macroeconomic theory is, in fact, incapable of addressing the questions that its proponents pose. More on that in a moment. Lepage seems to sense that inaptness in both his critique of Summers and in his embrace of the statistical methodology employed by the Bank of International Settlements. In his embrace of the latter, Lepage demonstrates that he is held captive by the disabling scheme of modern macro theory.
In this letter, I will explain why modern macro theory has so few answers to offer its practitioners. These theories focus on aggregate spending, when they should instead focus on the plans of economizing agents and how those agents achieve—or fail at achieving—coordination with one another. Modern macro theory gives pride of place to a few simple variables measured by aggregate spending. Those variables emerge from interactions among the myriad plans of economic agents which generate the observed spending variables.
Demonstrative and Plausible Reasoning
In her magisterial account of the history of economic theory, Mary Morgan explains that economists have no option but to reduce their sense of reality to models for the simple reason that economies are far too complex to be apprehended directly.1 But what kinds of models? This question has no easy answer. Intelligence and judgment must accompany any choice between models.
Economic theory began after theorists observed that recognizable patterns of human activity characterize social life. People do not live with the experience of being directed by a puppet master. Yet diverse human activities seem to fit into patterns. Economic theory arose through an effort to understand how generally coordinated patterns of activity emerge and evolve without a puppet master moving the strings.
The material of economics called for plausible reasoning, to recur to George Pólya’s treatment of plausible in relation to demonstrative reasoning.2 The early economists recognized that social reality was far too complex to be apprehended in its detailed entirety. They resorted to the construction of models to illuminate significant elements of social reality, while also recognizing that they would never capture the entirety of that reality. Late in the nineteenth century, some economists began to work with demonstrative reasoning. These later economists sought to offer mathematical proofs for a set of conditions under which the plans of all economic agents would mesh in a perfectly coordinated fashion.
While one can certainly appreciate the creativity that went into the demonstration of such abstract conditions, the elevation of demonstrability over plausibility that occurred during the twentieth century had worked its way into the mainstream of economic theory in the form of general equilibrium theory. Modern macro theory is organized around the model of dynamic and stochastic general equilibrium (DSGE). While developers of this model have received Nobel Prizes, the framework has also come under significant criticism since 2007. All the same, those criticisms have been limited to various theoretical margins, and have not sought to reconstruct economic theory in a fashion that would elevate plausibility over demonstrability.3
I have been involved in two projects which have pursued a program of plausible reasoning for thinking about systems of economic interaction.4 In one, I treated an economy as an ecology of interacting plans. In the other, Abigail Devereaux and I contrast the DSGE model with a model that is open-ended and evolutionary (OEE). We explain how the OEE framework can bring about a reconstruction of modern macro theory.
Axioms and Observations
Friedrich Nietzsche famously observed that we need multiple windows through which to observe the world. A model is a window onto the world. While a model allows us to see some things more clearly, it also prevents us from seeing other things. Equilibrium models allow us to see the world in terms of invariant structure. Yet structures change, and at varying speeds. Continents drift, although slowly when measured against human lifespans. In contrast, technologies change rapidly. Structure and change; pattern and process—both of these pairs of objects can provide a basis for theorizing about our observations regarding economic interaction.
Our ability to theorize depends on the analytical tools at our disposal. For a long time, those tools led us to theorize axiomatically in terms of equilibrium conditions. In recent years, tools have been developed that allow us to think systematically about systems that are continually evolving or changing.5 Agent-based computational modeling makes it possible to construct systems of interactions among participants, rather than simply positing the existence of such systems.6
In the axiomatic approach, economists reduce an economy to a few aggregate variables that are thought to describe the central features of an economic system. They proceed to formulate relationships among such aggregate variables as interest rates, money supply, investment, and rates of growth. This approach has been followed by mainstream theorists since the time of John Maynard Keynes. It is also the approach that Lepage takes, though in a somewhat different direction. The axiomatic approach, however, has a significant problem: system properties are built into the axioms, when they should rather be considered an emergent feature of interaction among the subjects who populate the model.
An alternative, the constructive approach, differentiates between primitive and derivative variables. Such standard variables as interest, money, investment, and growth are derivative variables. These are the results of prior economizing actions, which constitute the primitive variables of a system of economic interaction. Primitive variables are the objects at which people aim in forming their plans for economic action: they are the components of the ecology of plans that constitute a macroeconomy.
To recognize that economies have both primitive and derivative variables leads directly into an inquiry into the relation between the two types of variable. Orthodox macro theory posits a 1:1 correspondence between the two types of variable. It follows that any statement that could be made in terms of primitive variables could also be made in terms of derivative variables. This presumed correspondence is a feature of DSGE modeling. This standard manner of theorizing about economic systems has the virtue of being easily tractable. It also has the shortcoming of bypassing significant features of human population systems, features that might vitiate the conclusions that conventional macro models yield.
Modeling Human Population Systems
Consider a comparison involving two distinct human population systems where each contains the same number of members. One system is a parade and the other is a crowd of spectators leaving an arena. The parade is separated into a number of bands, floats, and equestrian units. Those members are readily reducible to a point on a map, and its movement can be traced. The parade is a system in equilibrium where disturbances must come from outside that system. The parade might be delayed if the engine on a float breaks down or if a flute player in the marching band collapses. These are exogenous shocks to a system that was presumed to have been in a state of equilibrium.
In contrast, the spectators leaving an arena cannot plausibly be reduced to some point of mass. Individuals move in different directions at different speeds. Both the parade and the spectators are orderly social configurations. A member of either group can operate sensibly inside their respective configurations. How that orderliness arises differs significantly between the two configurations. For the parade, orderliness is established by the parade marshal. If the marshal’s instructions are vague or even contradictory, the parade will perform at less than its potential.
Without doubt, it is analytically easier to model parades than to model crowds of pedestrians.7 All the same, the orthodox approach misconstrues the nature of the object it examines, which leads to misdiagnosis and ill-proffered advice. A crowd of spectators leaving an arena illustrates an ecology of plans. That ecology, moreover, is constituted through the plans of the various spectators. Their coordination is in some cases supplemented by traffic signals or sidewalk design, but most of the coordination among the pedestrians arises from a desire to arrive quickly at their destinations. Bumps and scuffles are mostly resolved through courtesy and mediation from within the crowd rather than through interjection from outside authority.
In Turtles, Termites, and Traffic Jams, Mitchel Resnick lamented the ubiquitous presence of a centralized mindset—a strong tendency among theorists to ascribe any observed orderly pattern to some leader who establishes that pattern.8 Modern macro theory reflects the same centralized mindset in its presumption that systemic order results from the actions of an ordering authority rather than residing mainly in the actions of participants.
In both the axiomatic approach and the constructive approach, there remains the question of what to do in the face of perceived poor performance. In the days of plausible reasoning, the bulk of economic theory was concerned with exploring the self-correcting features of economic systems.9 But within the DSGE framework, the economic system is assumed to have no ability to repair itself when exogenous shocks disrupt it. Most economists today appear to think that public policy will improve economic performance. Perhaps so, but belief in that possibility should be based on analysis of what political activity can accomplish within democracies.
Incorporating Politics
Economic theory began in feudal times, when government was the province of lords, and ordinary people tended to their own affairs. The advent of democracy, however, undermined the distinction between governors and governed. Conventional economic modeling still proceeds in a we–they framework, where expert mechanics work on engines that ordinary people cannot repair.10 However useful this framework might be for a king managing his kingdom or for an executive managing a corporation, it goes astray in democracies, where the rigid distinction between rulers and ruled vanishes or at least decays.
The orthodox scheme of thought—in which politics stands apart from economics—must give way to a scheme of entanglement between polity and economy.11 Politics is involved with commercial activity, and vice versa. Among other things, our common abstractions of markets and politics are outmoded. There is no such thing as a market economy whose performance properties are independent of politics. And there is also no such thing as some neutral political referee that enforces rules that people have previously agreed to.
Entanglement in political economy, it should be noted, is not something new. It was in play when economics proceeded through plausible reasoning and disappeared when demonstrative reasoning came to dominate economic theory. In 1933, Frank Knight explained that any economic system must address questions of resource allocation: what is produced, how it is to be produced, and how the product is to be distributed.12 Three years later, Harold Lasswell remarked that politics is centrally concerned with who gets what and how they get it.13 Neither Knight nor Lasswell referred to different domains of human activity in their treatments of economics and politics. Political and economic phenomena were entangled throughout a society and its myriad processes of human interaction.
When considering the financial crisis that began in 2007, economists who think in terms of DSGE modeling treat the system as having been hit by an exogenous shock and then inquire into the source of that shock. Some claim that the shock resulted from poorly working markets due to weak regulation. Others claim that excessive regulation was the culprit. An economist who sees an entangled political economy would point out that the present world is far removed from the early nineteenth-century world of classical liberalism. Current knowledge of human nature and contemporary process of political economy render unlikely any return to nineteenth-century modes of social living.
These days there is no such thing as a totally free market economy, and perhaps there never has been. According to the theory of markets, the suppliers of credit allocate credit to applicants according to calculations of profitability. A potential borrower will obtain credit only to the extent that a lender can be convinced to extend credit. But credit markets do not work this way. There exist politically supported regulations governing such matters as the distribution of loan portfolios across classes and categories of borrowers. These are formulated according to a logic of political as distinct from market profitability. Within democratic systems, moreover, politically supported regulations are not imposed by some ruling class that operates in splendid isolation from the remainder of society, but are emergent products of social interaction.
While we distinguish between economics and politics in our ordinary language, both types of activity are governed by the same analytical principles and human desires. There is a human population system that is constituted through myriad organizations that people establish, but those organizations interact within the same societal space. One of the results of such interaction is the social equivalent of plate tectonics. Commercial life proceeds differently in a system where lenders are able to deploy their assets according to a profitability calculus envisioned by the pure theory of a market economy than when credit is allocated inside a system governed by an admixture of political and profitability calculations.
Within democratic systems of political economy, the social world takes shape through transactions among enterprises of all sorts. Politically organized transactions have a triadic character, in contrast to the dyadic character envisioned by the pure theory of markets.14 Lepage is right to note the lingering quality of the 2007 crisis. That quality is not subject to mechanistic remedy. Periodic crises are an emergent feature of the societal tectonics that are part of a social system of deeply entangled political economy. What the future will bring remains to be discovered through contestation among societal participants within an analytical framework that is open-ended and evolutionary.
Richard Wagner
Henri Lepage replies:
I hope that Richard Wagner’s letter will be seen by many readers of Inference. It is a significant response and includes invaluable information about decisive developments in the field of economic theory and model construction.
I am not a professional academic. Throughout my career, I have remained a journalist interested in reviewing contemporary developments in economic thought from a libertarian perspective and with a focus on the Austrian school of economics. My initial intention—about four years ago—was to write a paper criticizing the foundations of modern macroeconomic models and explaining “why there is no such thing as macroeconomics.”15 As it turned out, the project was too ambitious. I subsequently changed course, refocusing on a new goal to debunk the responsibility of central-bank mainstream economics in the process leading to the great financial crisis and the painfully slow growth that has characterized its aftermath. I am now close to finishing this study and writing about it.
This background explains why I am so favorably disposed toward Wagner’s response. I am particularly excited about his new open-ended and evolutionary framework as a way to bring about a reconstruction of modern macro theory. This approach looks very promising and might be a suitable topic for a separate full-length analytical article.
During the last ten years, hopes for recovery have been dashed many times. Nonconventional monetary policies have, in fact, failed to bring about a return to stable growth. I do not think it too unreasonable to venture at this point that economic theory and ideas are now not only mistrusted, but also undergoing a process of disruption. The time has come for economic debates to shift back to fundamentals—as happened at the time of the Great Depression.
Wagner writes, “In his embrace [of the statistical methodology employed by the Bank of International Settlements, BIS], Lepage demonstrates that he is held captive by the disabling scheme of modern macro theory.” While it is true that the economists in Basel remain firmly ensconced within the core paradigm of mainstream contemporary economics, many of their papers contain criticisms that demonstrate their open-mindedness. The researchers at the BIS have also shown an unusual capacity to strike at the heart of the basic arguments upon which modern macroeconomic policies are built. This is all worth making more widely known to a public who are still in thrall to ongoing central-bank ideology and propaganda. It is a significant first step in a direction that I think is largely right.16
