The archive.

Between apriorism and a crude empiricism that would test every assumption in isolation, Machlup marks out a middle path for what 'verification' can mean in economics. He lowers the stakes deliberately: verification is not access to final truth but a disciplined comparison between what a theory implies and what inquiry discloses. His pivotal distinction separates particular historical propositions, checkable directly, from general hypotheses that yield conclusions only when joined to auxiliary assumptions about conditions and change. Fundamental postulates such as rational action need no independent sense-verification; they make conduct intelligible and are judged by the fruitfulness of the systems they support. Steering past both Misesian apriorism and Hutchison-style ultra-empiricism, and drawing on physics, Einstein, and Braithwaite, he holds that economics stays empirically disciplined while a hypothesis, at most, survives as not disconfirmed.

The hypothesis is confirmed if reasonable correspondence is found between the deduced and the observed, or more correctly, if no irreconcilable contradiction is found between the deduced and the observed.

A country may be poor with no promising resources to develop, or rich while sitting on vast undeveloped opportunities — 'how splendid to be rich and yet underdeveloped,' Machlup remarks, and the irony sets the tone for this 1957 map of development economics. Rather than treat 'economic development' as one measurable thing, he sorts the field's quarrels into disputes over definitions, objectives, recommendations, and theories, showing that total income, income per head, output per worker, and mass living standards can each move in opposite directions. Ends get confused with instruments: full employment, industrialization, and 'economic independence' harden into doctrines prized for their own sake. Behind the policy clashes he finds contested theories — autarky, infant-industry protection, balanced growth, terms-of-trade strategy — and asks of each exactly what exception to comparative advantage is being claimed.

The recommendation of labor-saving, capital-using investments in countries with abundant labor and scarce capital is a genuine paradox.

The slim volume in which Hicks reworked the demand theory of Value and Capital gets a patient, skeptical reading here. Machlup follows Hicks through weak ordering, the derivation of the law of demand, and the awkward Giffen case that requires an inferior good, a small substitution effect, and a large budget share at once. He sorts out the four consumer's surpluses Hicks had once collapsed into one, the First and Second Substitution Theorems, and the reciprocity of cross-effects. His verdict is measured: the revisions change neither policy nor prediction, but they honor Occam's razor by resting demand theory on fewer primitive terms, an important reconstruction, he grants, though difficult and joyless to read.

An important book is sometimes enjoyable to read. Unfortunately this cannot be said about this book.

Grant an inventor exclusive rights over a new process, and you plant a monopoly in an economy that otherwise prizes competition, the tension Machlup was asked to weigh when the Senate patent subcommittee commissioned this study. He traces the institution from Venice's 1474 law and England's Statute of Monopolies through the free-trade antipatent movement of 1850 to 1873, when the Netherlands actually repealed its patent law, and dissects the four classic defenses: natural property in ideas, monopoly as just reward, monopoly profit as incentive, and disclosure traded for temporary exclusivity. His conclusion is a famous refusal to conclude, that on present knowledge economics can pronounce the system neither a clear gain nor a clear loss to society.

To confuse an important invention with the patent that excludes people from using it is like confusing an important bridge with the tollgates that close it to many who might want to use it.

'Equilibrium,' in Machlup's hands, is a word doing too many jobs at once: a literal accounting balance, a methodological device for isolating cause, a description of durable historical states, and a smuggled value judgment. The confusion, he argues, comes from sliding among these senses unawares. To label a situation observed in the world an equilibrium commits what he calls the fallacy of misplaced concreteness, since the same facts fit rival models as equilibrium or disequilibrium; to build full employment or price stability into the definition is disguised politics. He carries the critique into international trade theory, praising Joan Robinson's relativism while faulting Nurkse, Ellsworth, and Kindleberger for defining the equilibrium exchange rate through the very policies they happen to favor.

Equilibrium is not a Good Thing, and disequilibrium is not a Bad Thing.

When an economist calls a country's troubles 'structural,' Machlup suspects he is more often excusing a policy than analyzing a problem, and here he puts the word on trial. He sorts its economic uses into three piles: clear senses worth keeping, such as the structure of production or econometric structural equations; vaguer senses better avoided, where structure means the slow-moving or the merely institutional; and a crypto-apologetic group in which structural imbalance props up cartels, price controls, exchange restrictions, and import barriers. Diagnosing a situation as structural disequilibrium, he argues, usually substitutes terminology for analysis. In a candid postscript he concedes that his least favorite examples happen to be policies he dislikes, an admission that his own politics may color which meanings he finds clear.

Structure, I am afraid, is often a weaselword used to avoid commitment to a definite and clear thought.

Neither instant copying nor a permanent monopoly serves an economy: some delay rewards the innovator who bore the sunk costs, but prompt competitive imitation is what turns one firm's advance into a general rise in productivity. From that premise Machlup builds a spare model of quasi-rents discounted over an expected imitation lag, and uses it to doubt that patents earn their keep. Long fixed terms, he argues, are a blunt instrument, too generous for cheap innovations, too weak for costly ones, and undercut anyway by circum-invention, substitutes, and discounting. The strongest case for a patent system, he suggests, is that it fosters a socially useful illusion of protection; the true optimum lag lies far closer to an innovator's natural head start than to sixteen years of statutory delay. The essay honors Fredrik Zeuthen.

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

Machlup gathered definitions of statics and dynamics from Comte through Patinkin and laid them side by side, only to find the terms pulling in incompatible directions: social order versus progress, stationary states versus growth, undated versus lagged variables. His mischievous observation is that economists tend to call their own theories dynamic and their rivals' static. Rather than legislate a single meaning, he builds a six-part typology that lets a theory like Schumpeter's qualify as dynamic under half a dozen headings at once, and shows how the supposed war between equilibrium and evolutionary economics dissolves once the labels are unpacked. His practical advice: retire the two words where possible in favor of exact ones, growth theory, sequence analysis, period analysis, trend analysis.

The trouble, as I see it, is not that the division of economic analysis into Statics and Dynamics makes no sense, but that it makes too many senses.

By the late 1950s American economists were split over whether rising consumer prices sprang from excess demand or from the wage and price decisions of unions and corporations, and Machlup thought both camps were arguing with blunt concepts. He rebuilds the vocabulary, separating autonomous demand from the induced and supportive kinds, and responsive cost increases from defensive and aggressive ones, then shows that cost-push cannot persist without monetary accommodation while demand-pull can stall against fully administered prices. Reading the postwar record through these lenses, he classes 1945 to 1952 as demand-pull and 1955 to 1959 as cost-push, with aggressive wage-push a likelier culprit than profit-push. His remedy is unfashionable: stable prices require that industries enjoying productivity gains be allowed to cut prices rather than pocket them.

A search of the learned literature would yield scores of definitions of inflation, differing from one another in essentials or in nuances.

Macroeconomic reasoning did not begin with Keynes: in this German-language essay Machlup runs the pedigree back through Quesnay, Ricardo, Cournot, Böhm-Bawerk, and Wicksell, then asks what really separates macro- from microtheory. Rejecting viewpoint and aggregation as the decisive test, he favors the absence of relative prices as the mark of a macro model, and insists that macro-relations like the consumption function rest on hidden micro-relations and unstable aggregations. He punctures the claims made for macrotheory's superiority, that it is more measurable, more dynamic, closer to policy, and warns against wringing causal conclusions from ex post identities. The firm of price theory, he reminds readers, is an ideal type, not a real enterprise. His verdict on the contest is deliberately anticlimactic: neither side wins, though microanalysis keeps a philosophical primacy.

Dennoch muß mein Urteil so lauten: kein Sieger und kein Unterlegener.

English translation: “Nevertheless my verdict must be as follows: neither victor nor vanquished.”

Einstein's definition of simultaneity, and Bridgman's attempt to turn it into a general rule that a concept means nothing more than the operations measuring it, set the stage for Machlup's defense of the unobservable. Physics itself, he shows, cannot do without freely invented constructs, from mass and inertial frames to Schrödinger's wave function, and neither can economics. Supply and demand curves, price, output, the industry are idealizations no statistic cleanly captures, muddied by heterogeneous goods, discounts, and quality changes; a model's weak fit with the data often indicts the data, not the model. Drawing on Menger's contrast of exact and empirical-realistic analysis, he defines a mental construct precisely and argues that a theory's concepts need not be realistic to be relevant, only its deduced consequences testable.

A mental construct is a concept designed for purposes of analytical reasoning that cannot be adequately defined or circumscribed in terms of observables or in terms of operations with recorded data derived from observation.

Does spending more on research simply buy more inventions? Machlup's answer is a patient no, or rather, much less than proportionately more. Inventive talent is scarce and its supply inelastic; drawing in extra researchers means paying rents to every incumbent and recruiting steadily less able hands, so marginal costs climb fast. Treating invention as an industry with a production function, he traces diminishing returns as more workers crowd a fixed stock of problems and knowledge, and catalogs ten reasons a swelling flow of raw ideas yields a rising share of rejects. The result is his four shrinkages, each thinning the passage from money spent to inventions actually put to work, a sober correction to any faith that funding alone accelerates technical progress.

These shrinkages are independent of one another; but they may add up with a vengeance.