Oskar Morgenstern’s conference contribution presents game theory as a method for analyzing strategic interdependence, with particular consequences for understanding power. Its progression—from basic concepts and scientific description through strategic choice to coalitions, bargaining, and voting—supports a deliberately qualified claim. Although the organizers supplied a title promising a general theory of power conflict, Morgenstern does not treat power as game theory’s original or exclusive object. Instead, exact descriptions of particular interactions should replace an encompassing but undefined vocabulary of power.
The opening methodological discussion makes utility the instructive comparison. Economics required a long conceptual development before von Neumann and Morgenstern could formulate axioms supporting numerical expected utility. Money is only one possible expression of a payoff; strategic analysis needs a broader account of what participants seek. Measurement itself presupposes conceptual work:
Die Theorie läuft der Messung voraus. Man muß zuerst eine Theorie aufstellen, dann folgt die Messung. Die Messung ist nicht von sich aus gegeben.
English translation: Theory precedes measurement. One must first establish a theory; measurement then follows. Measurement is not given in its own right.
Yet axiomatization exposes further difficulties rather than concluding inquiry. Morgenstern considers incomparable alternatives, partial orderings, and the problems of combining nonnumerical utility with expectations. The lesson for political science is that intuitive familiarity with power does not amount to possessing a usable theoretical concept. Substituting “influence” merely exchanges one undefined term for another.
This conceptual argument leads into a critique of description. Inventories of steel, workers, gold, or military forces do not adequately describe economic or political processes. What matters is how resources enter decisions, threats, and expectations. A weapon’s significance depends on its possible employment and on whether others believe a threat. Consequently, a model must capture the interaction of material capacities and anticipated conduct. It must resemble reality while remaining sufficiently simple to analyze and sufficiently complex to be illuminating. Appropriate mathematics may have to be developed alongside it, rather than borrowed indiscriminately from physics.
Game theory’s decisive departure from conventional economics concerns the distribution of control:
Daher wird die Spieltheorie eine Theorie des Konfliktes sowie der Zusammenarbeit sein, und sie will untersuchen, in welchen Fällen Zusammenarbeit wichtig ist und wie das, was durch die Zusammenarbeit erzielt wird, in einer stabilen Weise auf alle verteilt wird, die kooperiert haben.
English translation: Game theory will therefore be a theory of conflict as well as cooperation, and it seeks to investigate in which cases cooperation is important and how what cooperation achieves is distributed stably among all those who have cooperated.
Each participant controls only some variables determining the outcome. Morgenstern therefore rejects ordinary maximization as an adequate model of strategic interaction: unlike a decision-maker facing statistically describable natural conditions, a player confronts others deliberately pursuing their own objectives. His criticism of general-equilibrium theory is correspondingly structural. Adding constraints to a maximization problem cannot substitute for recognizing that control is divided among agents.
The minimax theorem exemplifies the alternative. Against an opponent with contrary aims, it supplies a way to minimize the worst possible outcome without an indefinite sequence of conjectures about the opponent’s conjectures. Mixed strategies protect players against having their conduct anticipated. Information, concealment, threats, and bluffing thus belong within the analysis rather than appearing as peripheral complications. Morgenstern then moves from two-person conflict to non-zero-sum interaction and coalitions, where cooperation creates gains but also generates disputes over their distribution.
The three-person majority game supplies his sharpest demonstration. Any two players can combine against the third; formally identical positions therefore permit exclusion in practice:
Das bedeutet, daß die Asymmetrie des Verhaltens die Symmetrie der Organisation zerstört.
English translation: This means that the asymmetry of behavior destroys the symmetry of organization.
Equal rules do not guarantee equitable outcomes. The example undermines the assumption that unrestricted interaction among formally equal participants must be fair. Further cases distinguish privilege from effective advantage and prejudice-based discrimination from exclusion arising through coalition structure. These are presented as results of simplified games, not as unrestricted empirical laws. Power can also reside in rules, while negotiation over which rules to adopt constitutes another game.
Distribution consequently requires a richer solution concept than a single maximizing outcome. Morgenstern emphasizes alternative allocations that do not dominate one another; their mutual nondomination gives a precise sense to the absence of power. His horse-market example makes the strategic point concrete. An unsuccessful bidder still raises the price paid by the stronger buyer. Once side payments become possible, the buyer may pay that rival to withdraw, while the seller may pay the rival to remain. Economic significance cannot therefore be inferred simply from who finally acquires the commodity.
The closing applications extend this approach to parliamentary voting and resource inequality. Shapley and Shubik’s work assigns values to participants through their roles in forming winning coalitions, rather than through headcounts alone. Morgenstern also distinguishes neutral games from power games, in which unequal resources affect available strategies. Repetition becomes especially difficult when each outcome transfers resources and changes the strategic possibilities of subsequent rounds.
The contribution’s relevance lies in this displacement of power as a general label by analyzable relations among rules, coalitions, information, and resources. Its confidence in exact methods remains bounded by an explicit acknowledgment of unfinished work:
Sie ist im gegenwärtigen Zustand weiß Gott keine Antwort auf alle Fragen der Ökonomie oder der Politikwissenschaft usw.
English translation: In its present state, it is certainly no answer to all the questions of economics or political science and so on.
Morgenstern’s concluding demand is to return from sharpened concepts to improved descriptions of reality. Game theory contributes not a finished definition of power, but a disciplined way to discover mechanisms and formulate questions that intuitive description alone could not reveal.
This work was divided into 4 sections when it entered the library's research corpus—an apparatus for search and citation, not necessarily the author's own table of contents. Each title opens its summary.
Put a question to this work; the Librarian answers from its 4 sections and cites the passage.
Ask the Librarian