Oskar Morgenstern
Oskar Morgenstern (1902–1977) was an Austrian-American economist whose collaboration with John von Neumann produced the foundational text of game theory. But to reduce him to "von Neumann's co-author" is to miss the entire point of what made their partnership productive — and what made it, in the end, asymmetrical.
Morgenstern brought to the collaboration something von Neumann lacked: a deep understanding of the conceptual problems that plagued neoclassical economics, and a conviction that those problems could not be solved by economic reasoning alone. He had been trained in the Vienna of the 1920s, where economics, philosophy, and the logic of science were not yet separated into disciplinary silos. His early work on economic forecasting — specifically his demonstration that perfect prediction is logically impossible in competitive markets because the act of prediction changes the behavior being predicted — was a direct precursor to the strategic interaction framework he would later formalize with von Neumann.
The Forecasting Problem and the Origins of Strategic Reasoning
In 1928, Morgenstern published a paper attacking the possibility of economic prediction. His argument was not empirical but logical: if an economist predicts that a stock will rise, and that prediction becomes known, rational agents will buy the stock immediately, causing the price to rise before the predicted time, falsifying the prediction. The forecaster and the forecasted are not independent; they are players in a game whose outcome depends on what each knows about the other's reasoning.
This was not merely a critique of forecasting technique. It was a critique of the entire neoclassical assumption that economic agents react to objective conditions rather than to expectations about others' reactions. Morgenstern understood, before game theory existed, that economic interaction is inherently strategic: your optimal choice depends on what you believe others believe you will choose. The infinite regress of mutual belief — I believe that you believe that I believe... — is not a philosophical curiosity but the defining structure of market behavior.
When Morgenstern met von Neumann at Princeton in 1938, he found a mathematician who had already proved the minimax theorem (1928) and who understood, with perfect clarity, that the problem Morgenstern was describing had a mathematical structure. Their collaboration produced Theory of Games and Economic Behavior (1944), a work that was neither pure mathematics nor pure economics but a new hybrid discipline: the formal study of interdependent decision-making.
The Asymmetry of the Partnership
The standard historical account treats Morgenstern as a junior partner — the economist who recognized the relevance of von Neumann's mathematics but contributed little to the formal apparatus. This account is false in both directions. Morgenstern did contribute to the formal structure: the concept of the characteristic function in cooperative game theory, the analysis of imputations and stable sets, and the insistence that the theory apply to all forms of social interaction, not merely to the zero-sum games that von Neumann had originally analyzed.
But the more important contribution was conceptual. Morgenstern forced von Neumann to address economics as it actually was — a domain of incomplete information, conflicting interests, and institutional complexity — rather than as a branch of applied mathematics. The book's insistence on applying game theory to cartels, markets, and voting systems was Morgenstern's agenda, not von Neumann's. Von Neumann would have been content to prove theorems; Morgenstern demanded that the theorems illuminate something.
Yet the asymmetry is real. The mathematical creativity was overwhelmingly von Neumann's. Morgenstern understood the problems; von Neumann built the machinery to solve them. The result was a book whose mathematical sections are von Neumann's and whose interpretive claims are Morgenstern's — and the two are not always consistent. The tension between formal rigor and economic relevance that runs through Theory of Games and Economic Behavior is the tension between two minds working at the edge of their respective competences.
Beyond the Book: Methodology and the Critique of Equilibrium
After 1944, Morgenstern's career diverged from von Neumann's. While von Neumann moved toward computing and military strategy, Morgenstern remained in economics, increasingly concerned with the methodological foundations of the field. He became a vocal critic of general equilibrium theory, arguing that the Arrow-Debreu framework assumed away the very problems — strategic interaction, information asymmetry, institutional constraint — that game theory had revealed to be central.
His later work on uncertainty, on the impossibility of economic prediction, and on the limits of mathematical economics placed him in an odd position: a co-founder of mathematical economics who became one of its most sophisticated critics. Morgenstern did not reject formalization; he rejected the confusion of formalization with understanding. A model that is mathematically elegant but economically empty is, in his view, worse than no model at all, because it creates the illusion of knowledge.
This critique has been largely forgotten, overshadowed by the triumph of game theory in economics after Nash. But it is worth recovering. Morgenstern's skepticism about equilibrium concepts — his insistence that real economies are disequilibrium systems, constantly disturbed by innovation, information arrival, and strategic surprise — anticipates much of the work in complexity economics and agent-based modeling that emerged decades later. The economist who helped formalize strategic rationality ended his career arguing that formal rationality is not enough.
The standard history of game theory is written as the progressive refinement of equilibrium concepts: von Neumann-Morgenstern stable sets, then Nash equilibrium, then refinements, then evolutionary dynamics. This history treats the 1944 book as a starting point and everything after as improvement. But Morgenstern himself would have rejected this teleology. He understood that the project of formalizing strategic interaction was not a ladder to be climbed but a map to be questioned — and that the most important question was not 'what is the equilibrium?' but 'what does the equilibrium obscure?' The fact that this question has been systematically ignored by the field he helped create is not a sign of progress. It is a sign that the field has forgotten why it started.