Fritz Machlup, Concepts of Competition and Monopoly—Discussion (1955)

Fritz Machlup’s contribution to a joint conference discussion responds to papers by Clark and Weintraub on competition, monopoly, and economic theory. It moves from the social costs of monopoly and the incentives for innovation to questions of analytical method: equilibrium, dynamics, oligopolistic indeterminacy, product differentiation, and profit maximization. Its connecting concern is the danger of treating either reassuring empirical estimates or broad theoretical labels as substitutes for carefully specified arguments. Machlup defends vigorous competition while insisting that economic models be judged by the problems they can illuminate, not by their apparent realism or disciplinary prestige.

The opening challenges recent assurances that monopoly poses little danger: competition supposedly remains effective, large firms finance technological progress and widen consumer choice, countervailing power offsets monopoly, and monopoly’s annual welfare cost amounts to only $1.50 per person. Machlup presents these claims as grounds for complacency about antitrust enforcement. He welcomes Clark’s and Weintraub’s resistance, particularly Clark’s refusal to equate bilateral monopoly with competition. Bargaining between powerful parties does not ensure a competitive outcome, and the capacity of large enterprises to fund research does not settle whether monopoly serves society.

Innovation supplies his first substantive distinction. Society benefits both from generating new techniques and from spreading their advantages through imitation. Competitive erosion of an innovator’s profits renews the incentive to innovate, whereas excessive protection postpones diffusion. Quoting his earlier work, Machlup states:

To buy innovation by paying with unnecessarily long delays of imitation is a poor bargain for society to make.

His proposed criterion for patent protection turns on the difference between expected and actual protection. Innovators who overestimate their head start may undertake more innovation without imitation actually being delayed. Conversely, protection stronger than innovators believe it to be imposes diffusion costs without a corresponding incentive benefit. The argument therefore evaluates patents through the relationship between expectations, incentives, and realized delays, rather than treating monopoly protection as an undifferentiated prerequisite of progress.

Machlup next questions welfare calculations that identify monopoly power through unusually high profit rates:

Many writers on the subject have insisted that monopoly is reflected less in increased profit rates than in increased or inflated costs.

Monopoly rents can disappear into inflated payments for productive services, inefficient resource use, or capitalized asset values that enlarge the denominator of the measured rate of return. High management and labor costs may likewise absorb rents. Tariffs, patents, building codes, and mineral monopolies furnish examples of barriers behind which excessive costs can persist. His criticism is directed at what a profit-based estimate excludes: modest recorded profits need not demonstrate modest social losses.

Despite these policy implications, Machlup locates the papers’ chief value in their theoretical contributions. His praise of Clark’s dynamic appraisal leads into a demand for greater terminological precision. Equilibrium is a tool for explaining change, not the opposite of growth. Both static models that disregard time lags and dynamic models that describe temporal sequences can employ equilibrium concepts. The relevant contrast is consequently between particular static analyses of price and output and theories of growth, not between equilibrium and development as such.

A selling campaign illustrates why temporal sequencing sometimes matters: identical selling efforts can produce different outcomes depending on the order of actions and the intervals between them. But the existence of these differences does not establish their importance for every analytical question.

Dynamic theory is better only where it is needed; where the analytic job can be done with the simpler, less realistic, static theory, the latter is preferable.

This defense of selective simplification also governs his objection to making indeterminateness a defining feature of fully dynamic theory:

I should prefer to use language which does not mix the dynamic elements, the uncertainty elements, and the indeterminateness, since all these are different things that ought to have separate names.

Oligopoly may require all three, but their conjunction does not erase their conceptual differences. Machlup agrees with Clark that the interval between a move and a countermove is central to oligopolistic competition. His objection concerns the vocabulary used to identify mechanisms, not the importance of those mechanisms.

The same attention to scope shapes his defense of the tangency theorem. The objection that firms may lower prices to forestall entry concerns oligopolistic behavior; Machlup understands the theorem as addressing a polypolistic seller too small to anticipate competitive responses. Product differentiation similarly cannot be assigned one competitive effect: quality rivalry may intensify competition, while standardization can facilitate either price competition or collusion. He resists attributing a single position to theory when only some theorists hold it.

The concluding methodological exchange challenges Weintraub’s objection that subjective profit maximization merely says businessmen do whatever they think right. Machlup argues that marginal analysis can predict characteristic responses to a change, such as a rise in fuel prices, without reconstructing every belief underlying the initial position:

The point is that, in order to predict or explain the type of adjustment, we do not have to know the subjective whims and guesses that may have influenced his initial position.

The contribution’s lasting relevance lies in this combination of competition policy and methodological restraint. Monopoly costs require more than profit accounting, and theoretical usefulness requires neither exhaustive realism nor indiscriminate dynamization. Machlup ends by proposing reciprocal reading and discussion with Weintraub: conceptual disagreement should produce closer examination of arguments, not sweeping verdicts about what theory must say.