4,366 works (including 487 books, 3,476 articles and 400 historical documents), 150 years of economic thought. Each one summarized and searchable.
'Autonomous' and 'induced' name positions in a causal story, not fixed labels stamped on the current account, the capital account, or official reserves. Against the habit of reading causation straight off the balance-of-payments table, Machlup insists that double-entry identities guarantee offsetting balances but explain nothing: a trade surplus is itself a capital export, and a current-account deficit is the logical correlative of a capital-account surplus rather than its cause. Using an oil-price shock to show that an importing country may cut consumption, raise exports, borrow, or draw down reserves, he weighs five rival doctrines of autonomy and ties each to a specific exchange-rate regime. His verdict is disciplinary: which flow dominates an episode must be established by historical and theoretical argument, never inferred from account headings alone.
Economists are prone to consider assumptions as almost perfect substitutes for knowledge, or perfect antidotes for ignorance.
Expensive to create, verify, and teach yet nearly free to reuse, knowledge resists efficient pricing—a tension that runs through this survey of information economics and human capital. Writing deliberately without a line of algebra, Machlup guides the reader through markets riddled with asymmetric information, from Akerlof's lemons to adverse selection and moral hazard, through public goods and the free-rider problem, and back to the Mises-Hayek socialist calculation debate and the dispersed knowledge that no central board can gather. He accepts a weak rational expectations while rejecting the strong version as granting agents superhuman powers, and narrows human capital to investments built into persons alone, distinct from tools and disembodied knowledge. The third volume of the Knowledge project, published after his death, completes it.
I have set for myself the task of writing without a single line of algebra, even where this constraint should make it impossible to give a proper presentation of an author’s ideas.