Werner Sombart called capitalism and double-entry bookkeeping form and content, and the evidence has run against him since 1949

Few claims about business technique have travelled as far on as little evidence as Werner Sombart's about bookkeeping. In Der moderne Kapitalismus, first published in 1902 and expanded later, Sombart argued that capitalism and double-entry bookkeeping were inseparable, standing to each other as form to content, and that it was simply impossible to imagine capitalism without it.¹² He compared it to the achievements of Galileo and Newton, calling it a cosmos built on mechanistic thought.¹ Translations of these passages differ, but the thesis is not in doubt, and it is easy to see why: it gave capitalism a technology and therefore a date, which a story about the spirit of an age otherwise lacks.

The chronology is the first difficulty. Luca Pacioli's Summa de arithmetica, printed in Venice in 1494, contains the earliest printed systematic description of the method, which Pacioli called the method of Venice and documented rather than invented; Venetian ledgers from 1407 already show it in use.³ Alan Sangster's analysis of a ledger kept by Florentine moneychanger-bankers in 1211 pushes the origin back further still, to northern Italian moneychanger-bankers of the twelfth century.⁴ A technique available three centuries before Pacioli described it, and six before anything most historians call capitalism, is a poor candidate for a cause. That is not fatal, since a precondition may lie dormant, but it shifts the burden onto the mechanism.

Basil Yamey, an economist at the London School of Economics, took up that burden in 1949 and did not put it down for four decades. His test was the obvious one: if double-entry mattered because it let merchants calculate profit, then merchants using it should have calculated profit. Yamey looked for evidence of regular balancing of the books and did not find it. He cited Richard Dafforne's manual of 1651 on the occasions when a balance was called for: the books were full, the merchant had ceased trading, or the owner had died. Assets were rarely revalued. His judgement was that there was little evidence of a careful calculation and analysis of profit, because most merchants were so intimately concerned with their own affairs that they had no need of elaborate calculation.⁵ He returned to the argument in 1964 and after, holding that Sombart had read too much economic significance into the method.⁶

If the method was not for calculating profit, what was it for? The answer that has emerged is about agency rather than arithmetic. A principal who cannot watch an agent needs a record the agent cannot quietly alter, and a system in which every entry appears twice and the whole must balance is precisely that: an error or a theft breaks the identity and announces itself. Sangster puts the business need as controlling credit relationships and demonstrating accountability rather than any desire to produce a balance sheet, and his survey of diffusion before 1800 finds that few merchants used double-entry to establish overall profit or wealth, and that those who did were accountable to others, such as distant partners and factors.⁴⁷ The method is a governance technology that can compute a profit, not a profit calculator that improves governance.

Two complications belong on the table rather than buried. Yamey's evidence came largely from printed bookkeeping manuals rather than surviving merchant ledgers, few of which remain, so he partly tested what authors recommended rather than what merchants did; and his definition of double-entry drifted, at times requiring profit and loss accounts and balance sheets on twentieth-century reporting lines, which makes a finding of absence easier to reach.⁷ Critics add that he asked what an individual trader needed for their own decisions, which is not the same question as whether a society was changing.¹ Both are fair, and neither restores the profit-calculation story, because both concern whether Yamey proved too much rather than whether merchants computed returns on capital.

The most useful resolution came from sociology. Bruce Carruthers and Wendy Espeland argued in 1991 that double-entry can only be understood if its rhetorical and technical sides are taken together: accounting is not merely computation but an attempt to convince an audience that a venture is legitimate. Surveying practice from the Middle Ages to the nineteenth century, they found slow adoption, books often poorly kept, and substantially symbolic motives.⁸ Yamey thought that conclusion highly improbable.⁸ He was right to resist it as a replacement and wrong to resist it as a complement, because a record built so that an agent cannot alter it without detection is exactly the record a stranger can be asked to trust. Legitimacy is not an alternative to control; it is what control looks like from outside the firm.

Which brings the argument to where the textbook version misrepresents the original. Sombart's emphasis was on separating the firm's capital from the owner's person, and that has survived rather well: it is hard to run a joint-stock company without it. What passed into accounting textbooks and popular histories instead was a claim about rational profit calculation, presented with more confidence and supported by less. The weaker link got transmitted. Work on the Dutch East India Company makes the point from the other direction, concluding that double-entry was not a necessary condition for Dutch capitalism.⁹

This matters beyond history, because a measurement system's purpose can be read off its design. Double-entry was built for completeness and verifiability: historical cost, every entry twice, the whole constrained to balance. That makes it good at answering what was spent, received and owed, and awkward at what something is now worth, because the conservatism that defeats a dishonest clerk also defeats revaluation. It is not a quirk of modern standards but the residue of the original purpose, and it is why accounting's long quarrels, over fair value against historical cost, over whether research spending or a brand belongs on a balance sheet, keep taking the same shape: they are attempts to get valuation out of a system designed for stewardship.

The obvious objection is that this is antiquarian. Firms today calculate profit continuously, so whatever the ledgers were for in 1211 can hardly bind a modern accounting department. The objection understates how much of the architecture survived. A company's accounts are still built on the double-entry identity, still anchored in transactions that happened, and still audited for whether the records faithfully represent those transactions rather than whether the enterprise is correctly valued. The asymmetry that follows is familiar to anyone who has compared a firm's book value with its market value.

The account suggests a test. If stewardship rather than calculation is the explanation, accounts should be most elaborate where an agency relationship is sharpest and thinnest where none exists. Audited statements to absent shareholders, trustee accounts and the books of a manager reporting to a distant partner should be the developed cases; an owner-managed business, which has nobody to convince, should be the least. That is roughly the pattern then and now. Sombart was looking for the invention that made a new economy thinkable. The ledgers describe something less romantic and more durable: a technique for making it safe to let someone else handle the money.

Notes

1Werner Sombart, Der moderne Kapitalismus (first published 1902; the much-quoted passages appear in the expanded later edition of 1916-1927), chapter on the birth of the capitalist enterprise, for the claim that capitalism and double-entry bookkeeping are inseparable as form to content, for the statement that it is impossible to imagine capitalism without double-entry bookkeeping, and for the comparison with the systems of Galileo and Newton and the description of the method as a cosmos built on mechanistic thought. The original was not opened and translations of these passages differ materially between sources, so no single rendering is quoted as definitive here. The criticism that Yamey addressed individual decision-making rather than social change is reported in the same literature. Kenneth S. Most's translation of the relevant chapter

2Jacques Richard and others, "Accounting and the birth of the notion of capitalism," Critical Perspectives on Accounting, for the rendering of Sombart's thesis that capitalism and double-entry bookkeeping are absolutely indissociable and that their relationship to each other is that of form to content, and for the placement of the claim in the wider literature on accounting and the concept of capitalism. Publisher page and an open repository record

3Luca Pacioli, Summa de arithmetica, geometria, proportioni et proportionalita (Venice, 1494), for the earliest printed systematic description of double-entry bookkeeping, for Pacioli's own name for it as the method of Venice, for the memorandum, journal and ledger structure he sets out, and for the fact that he documented an existing merchant practice rather than inventing it. Surviving Venetian ledgers from 1407 already show the method in use. Sources differ on the length of the bookkeeping treatise within the Summa and on Pacioli's birth year, neither of which is relied on here. Institute of Chartered Accountants in England and Wales on Pacioli, the same library's survey of the earliest books on bookkeeping and a catalogue account of the Summa

4Alan Sangster, "The emergence of double entry bookkeeping," The Economic History Review 78, no. 2, pp. 499-528, DOI 10.1111/ehr.13358, for the critical analysis of entries in a ledger of Florentine moneychanger-bankers of 1211, for the attribution of the method's origin to northern Italian moneychanger-bankers in the twelfth century, and for the framing of the business need as the control of credit relationships and the demonstration of accountability rather than a desire to produce a balance sheet. Sources differ on whether to date the article 2024, when it appeared online, or 2025, the journal volume year. Not opened; the abstract and citing summaries were returned in search results. Publisher page, the University of Aberdeen research record and an Economic History Society summary

5Basil S. Yamey, "Scientific Bookkeeping and the Rise of Capitalism," The Economic History Review 1, no. 2-3 (1949), pp. 99-113, for the rejection of Sombart's claim that double-entry bookkeeping was a precondition or important stimulus for modern capitalism, for the search for evidence of regular balancing and the finding that regularity was not a feature of the practice, for the citation of Richard Dafforne's 1651 manual on the occasions requiring a balance, for the observation that revaluation of assets was rarely carried out, for the judgement that there was little evidence of a careful calculation and analysis of profit, and for the suggestion that the majority of merchants were probably so intimately concerned with the details of their own business affairs. Not opened; the quoted phrases were returned in search results quoting the article, and page ranges differ between secondary sources. Publisher page for the 1949 article

6Basil S. Yamey, "Accounting and the Rise of Capitalism: Further Notes on a Theme by Sombart," and his 1964 paper in the Journal of Accounting Research, for the continuation of the argument and for the judgement that Sombart was mistaken in reading too much economic significance into double-entry bookkeeping. Secondary sources give conflicting page citations for the 1964 paper, so none is given here. Bibliographic record of the Further Notes paper and a biographical record of Yamey, for his position at the London School of Economics

7Alan Sangster, "The diffusion of double entry bookkeeping before 1800: establishing a new research agenda," for the finding that merchants used double-entry principally for control over debt and over distant agents, factors and partners, that few used it to identify overall profit or wealth, and that those who did so were accountable to others; and for the observations that Yamey's conclusions rested largely on printed bookkeeping manuals rather than surviving account books, and that his definition of double-entry shifted over his career towards one conflating bookkeeping with financial accounting. Open-access copy at the University of Aberdeen repository

8Bruce G. Carruthers and Wendy Nelson Espeland, "Accounting for Rationality: Double-Entry Bookkeeping and the Rhetoric of Economic Rationality," American Journal of Sociology 97, no. 1 (1991), pp. 31-69, DOI 10.1086/229739, for the argument that the significance of double-entry bookkeeping can be appreciated only if its rhetorical and technical aspects are considered together, for the claim that accounting is an attempt to convince an audience of the legitimacy of business ventures, for the survey of practice from the Middle Ages to the nineteenth century, and for the reported findings of slow adoption, poorly kept books and substantially symbolic motives. Yamey's description of the symbolic-adoption conclusion as highly improbable is reported in secondary summaries of the exchange rather than taken from his own text, which could not be reached. Bibliographic record with DOI and a second record of the article

9For the conclusion that double-entry bookkeeping was not a necessary condition for Dutch capitalism, drawn from study of the Dutch East India Company's accounting for social capital between 1602 and 1623. Publisher page