Pathologies · Chapter 8

When does inequality break a regime?

History’s ledger case is France in 1789, and it does not say what the slogan says. Britain owed three times as much against its output and taxed nearly twice as much of it, and nothing happened there. What France could not do was collect from the wealth it protected, and the reckoning arrived the winter bread reached 14½ sous.

In this chapter

The phrase turns up in a caption under a photograph of a yacht, and in a comment under a story about a pay ratio, and eventually at a dinner table, said half as a joke by someone who is not entirely joking. Eat the rich. It is old, and it is elastic, and underneath it sits a claim that is genuinely testable: that when the gap gets wide enough, the thing tips over. People who say it usually have a picture in mind, and the picture is France. Whether the rich got rich off everyone else is a question about margins, and it has its own home in the frameworks volume’s chapter on whether the economy is zero-sum. This chapter takes the second half, the half the slogan is really asking. How does it end? And it goes to the case the slogan is borrowing, because that case has account books.

They are damaged account books. There is no French Treasury ledger to look up. Every number in this chapter is a reconstruction, assembled from what survived by a named historian, which is why every figure below is attributed to the person who built it. It also turns out to matter more than a caveat usually does, because the two economic historians who built the most careful reconstructions of the French crown’s position both concluded that the thing everyone blames did not cause what everyone says it caused.

The crown’s position, as far as anyone can rebuild it

The usable document is an account of the projected budget for 1788, published by Brienne, and the modern route to it runs through Frédéric Braesch, who summarized it in 1936, and then through Michael Bordo and Eugene White, who read Braesch. On their reading the crown expected revenues of about 472 million livres and expenditures of about 565 million, of which 260 million were interest payments, leaving a deficit of about 93 million that the finance minister planned to cover by borrowing more. A state spending a fifth more than it takes in, with roughly half of what it spends going to its creditors, is a state in trouble. Figure 8.1 lays out the position and then does the thing that changes its meaning, which is to put the same ratios beside Britain.

A CROWN THAT COULD NOT COLLECT, AND THE HISTORIAN BEHIND EVERY ROW no Treasury ledger survives; each figure below is a reconstruction, named THE PROJECTED BUDGET FOR 1788 expected revenues about 472 million livres expenditures about 565 million of which interest payments about 260 million deficit, to be covered by new borrowing about 93 million Bordo and White, from Braesch’s summary of Brienne’s published account THE SAME RATIOS, THE SAME YEAR, TWO CROWNS FRANCE 1788 BRITAIN 1788 debt service, as a share of tax revenues 61.9% 56.1% taxes collected, as a share of national output 6.8% 12.4% government debt, as a share of national output 55.6% 181.8% all six figures: Weir, Table 1 · Britain owed more, collected more, and did not break WHAT THE CROWN COULD NOT REACH taxes on income without regard to status or privilege collected 12% of revenues in 1789, with, in the source’s words, “the persisting irregularities in assessment” the clergy’s tax bill averaged 10 million livres a year, under 2% of state revenues, while the church owned 10–15% of the property in France nobles were exempt from the taille, but partially: a noble working his land with servants was exempt, while a sharecropping tenant’s share stayed taxable first two rows: Sargent and Velde · third row: White
Figure 8.1 The crown’s 1788 position, with the reconstructor of every row named, because there is no ledger to consult and each of these figures is somebody’s rebuild. The middle block is where the diagnosis turns. Read the first row alone and France looks crushed: 61.9 percent of its tax revenues went to servicing its debt. Read the next two and the diagnosis inverts. Britain, in the same year, owed more than three times as much against its national output and was paying a slightly smaller share of its taxes to do it, because it collected nearly twice as much of its economy in tax. France’s debt was not large. Its tax base was small. The bottom block is why it was small, at the precision the record supports, which is coarser than the slogan’s. What it documents is a patchwork of privileges, waivers and negotiated lump sums, in which a noble’s liability turned on how his land happened to be farmed, rather than a clean exemption of the rich. The church owned something between 10 and 15 percent of the property in France and supplied under 2 percent of the state’s revenues, by a payment it conceded rather than a tax it owed. Budget: Michael D. Bordo and Eugene N. White, “A Tale of Two Currencies,” NBER Working Paper 3517 (1990), p. 10, as they state it, from Frédéric Braesch, Finances et monnaie révolutionnaires (Paris, 1936), summarizing Brienne’s Compte rendu au Roi of March 1788; figures given as the paper’s own rounded text, not to the decimal. Ratios: David R. Weir, “Tontines, Public Finance, and Revolution in France and England, 1688–1789,” Journal of Economic History 49(1), 1989, Table 1, p. 98 — debt service here includes redemption, not interest alone. Exemptions: Thomas J. Sargent and François R. Velde, “Macroeconomic Features of the French Revolution,” Journal of Political Economy 103(3), 1995, pp. 484–485; Eugene N. White, “France and the Failure to Modernize Macroeconomic Institutions,” Rutgers working paper 1999-04, pp. 8 and 16. Retrieved 2026-07-17.

The comparison with Britain is not a curiosity. It is the reason this chapter cannot give the slogan what it wants, and the historians who assembled these figures say so themselves, in sentences that cut against the story the case is usually made to carry. Weir’s is flat: “The debt burden in 1788 cannot explain why France had a revolution. The past history of the debt burden itself cannot either.” He points at his own Figure 1, which tracks debt interest as a share of taxes across a century, and notes that France apparently mirrored Britain through the same wars while keeping a consistently lower interest burden on its revenues, on data he cautions are weakest before 1726. Sargent and Velde reach the same wall from the other side. France, on the evidence they assemble, had somehow solved the sovereign debt problem: bond prices in the late 1780s were not especially depressed, and the Revolution would strain for eight years to honour the king’s debts before defaulting. Their diagnosis is not about the size of anything. “Up to 1789, France’s fiscal arrangements had evolved unevenly, and the ability to adjust taxes did not match the king’s plans to service his debts.” France, they write, lived with an array of institutions whose purpose was not only to fund the king but to constrain him, and “These constraints impinged particularly when it became desirable to raise taxes to service debts.”

That is the mechanism, and it is narrower than the slogan. The crown was not bankrupted by its debt. It was trapped between a debt it intended to honour and a tax system it was not permitted to fix, and the trap closed at precisely the moment it needed to raise money. White, who spent a career in these accounts, put the summary in one line: “Privilege shaped French society and privilege shaped the fiscal system, with vested interests it created blocking reform efforts.” The privileges were not merely unfair. They were load-bearing. The vingtième was meant to be a five percent tax on everyone’s income; the clergy won exemption from it and the provincial estates paid lump sums instead. Each escape was a negotiation, each negotiation created a party with an interest in the escape, and by 1788 the crown had to convene the Estates General to get a tax through, which meant assembling in one room every group that had spent a century learning how to stop one.

It is worth being exact about Necker here, because he is the figure the story usually leans on. His 1781 Compte rendu claimed that his reforms had restored balance in the “ordinary” budget, and the claim is famously misleading. But the reason is more interesting than fraud. As Velde and Weir put it, that ordinary budget “was a projection of a peacetime budget to follow the American War and ostensibly included all charges incurred by the war as of early 1781, but certainly not the war bills that arrived later.” The verdict of fabrication is Marcel Marion’s, and Velde and Weir immediately record that Necker has found new defenders who find his estimates convincing. So the Compte rendu is not usable as a fiscal source, and it is not used as one here. What it is evidence of is the opacity, which Bordo and White state precisely: the monarchy was able to deceive the public “because government finance was not open to parliamentary inspection as in Britain.” A state whose accounts nobody may audit is a state that can defer the reckoning, and deferral is how a fiscal problem becomes a political one.

The winter the bread went up

A fiscal trap is not a revolution. What turned one into the other in the summer of 1789 was the price of bread, and here too the record belongs to a reconstructor: George Rudé, who rebuilt the Paris loaf price day by day out of the journal of the bookseller Hardy and the archives of the trade administration. The series is his, not Labrousse’s, whose own reconstruction is a different thing at a different scale, the price of wheat across the généralités of France, on which the rise to the July 1789 peak was, in Labrousse’s own words, “only 50 per cent compared with the cyclical minimum of 1786.” Figure 8.2 is Rudé’s Paris series.

THE PRICE OF THE 4-LB LOAF IN PARIS, OCTOBER 1788 TO SEPTEMBER 1789 sous; a step, because the record is of dated changes 8 10 12 14 8–9 sous before the crisis THE BASTILLE FALLS, 14 JULY the price does not move for another week 14½ sous from 1 February, and there it stays 12 13 14 13½ 12 O N D J F M A M J J A S 1788 1789 WHAT THAT WAS AS A SHARE OF A WAGE · the weaker record, and it says so a builder’s labourer spent 80% of his effective earnings on bread between February and July, and bought the same bread with 67% after the August cut; a journeyman, 48% then 40% Rudé. “Effective” earnings allow for 111 unpaid feast days a year, a figure he takes from Jaffé; and nearly all his wages, by his own note, are 1790 data, though he judges 1789 very close
Figure 8.2 The Paris loaf, reconstructed by George Rudé from the bookseller Hardy’s journal and the trade archives. A four-pound loaf had sat between 8 and 9 sous for years. It reached 14½ sous on 1 February 1789 and then did something more telling than spiking: it stopped there and stayed, month after month, through the spring in which the Estates General met and the summer in which the Bastille fell. The vertical rule marks 14 July, and the line beside it is flat. Rudé writes that the price was “to remain at this level until after the fall of the Bastille,” and the first cut came on 22 July, eight days later, with a second after demonstrations outside the Hôtel de Ville on 8 August. Bread did not spike and topple a regime. It went up by about two thirds and then sat on the city for half a year, which is the difference between a shock and a condition. The lower block is the same story as a share of what a man earned, and it is the weaker half of this plate by its author’s own account: those shares are of “effective” earnings, a construct of Rudé’s that deducts 111 unpaid feast days a year on a figure he borrows from Jaffé, and nearly all of his wage data, as his footnote concedes, are for 1790 rather than 1789, though he adds that what evidence exists for 1789 suggests the two years differ very little. George Rudé, The Crowd in the French Revolution (Clarendon Press, 1959), pp. 25, 33, 67, 68 and Appendix VII, p. 251; read via the Internet Archive’s search-inside service, which returns the passages verbatim. The wheat-price series for France as a whole is Ernest Labrousse’s and is a different measurement at a different scale: La crise de l’économie française à la fin de l’Ancien régime et au début de la Révolution (2e édition, Presses universitaires de France), on which the rise to the July 1789 peak was “only 50 per cent compared with the cyclical minimum of 1786.” Retrieved 2026-07-17.

Put the two figures in the same room and the sequence is visible. A crown that cannot raise taxes without convening the people who have spent a century blocking taxes convenes them, in 1789, because the alternative is the default it has been straining to avoid. It convenes them into a city where a labourer has been handing over four fifths of his effective earnings for bread since February. Neither fact causes the other. The fiscal trap did not raise the price of bread, which was a matter of a bad harvest and a brutal winter; the bread did not create the fiscal trap, which had been assembling for generations. What they did was arrive together, and the arrival put a desperate population and a paralysed state in the same place at the same time, with a newly summoned assembly sitting between them, discovering that it was the only body in France that could legally do anything.

Conditions, not a law

So what does the case actually license? Not a prediction, and the modesty is not politeness. It is arithmetic. This is one revolution. Anything of the form “when inequality passes here, regimes fall” requires a set of cases with a line through them, and a set of one has no line. What the case does yield is a list of conditions that were jointly present, each of which the record documents, and each of which could be looked for elsewhere: a state whose obligations were rising while its ability to collect was fixed by privilege it could not revoke; accounts no independent body could inspect, which let the problem be deferred past the point of easy repair; a subsistence squeeze severe enough and long enough to make a large population available; and an institution that had to be summoned to solve the first problem and turned out to be the instrument for the rest. Remove the third and you have a fiscal crisis, which France had survived repeatedly. Remove the first and you have a bread riot, which Paris had survived repeatedly. What was new in 1789 was the conjunction.

Notice what is not on that list. Not the gap between rich and poor, which no figure in this chapter measures and which none of these reconstructors set out to measure. Not the debt, which Britain carried more of. Not the wealth of the aristocracy as such. What appears on the list, every time, is that wealth’s relationship to the state’s capacity to reach it. The French nobility and the French church were not brought down because they were rich. What the record documents is narrower: their privileges were the exact shape of the hole in the crown’s revenue, and the crown could not close that hole without dismantling arrangements it was built on. That is a much more specific condition than inequality, and it is the one the reconstructions record.

Inequality inevitably leads to revolution.

Oversimplified Moderate confidence

The claim has a true kernel, and 1789 is it: a society organized around privilege did come apart, and the privileges were not incidental to how it came apart. They were the mechanism. That is what keeps this ruling off backwards, and anyone who answers that the distribution of wealth had nothing to do with the fall of the French monarchy has to explain why the crown could raise only 6.8 percent of national output in tax while the church supplied under 2 percent of revenues on 10 to 15 percent of the property. The word that fails is “inevitably,” and it fails on the evidence assembled by the people who built this chapter’s figures. Britain in the same year owed more than three times as much against its output, paid a comparable share of its taxes to service it, and had no revolution. Weir, whose table that is, writes that the debt burden in 1788 cannot explain why France had a revolution. What the record supports is narrower: a particular arrangement came apart, in which the wealth the state protected was the wealth it was structurally unable to tax, at a moment when it had to. Inequality is the raw material of that story. It is not the mechanism, and the case cannot be run in reverse to forecast anything.

Sources
  • The fiscal position: Michael D. Bordo and Eugene N. White, “A Tale of Two Currencies,” NBER Working Paper 3517 (1990), p. 10, reading Braesch’s summary of Brienne’s account for 1788; and their note that the monarchy could deceive the public “because government finance was not open to parliamentary inspection as in Britain.” All of it is reconstruction: no Treasury ledger survives, and the figures are given as their own text states them rather than to the decimal.
  • The counter-case, from the same authorities: David R. Weir, “Tontines, Public Finance, and Revolution in France and England,” Journal of Economic History 49(1), 1989, Table 1, p. 98 (France and Britain 1788: debt service 61.9 against 56.1 percent of tax revenues; taxes 6.8 against 12.4 percent of output; debt 55.6 against 181.8 percent of output), and p. 101: “The debt burden in 1788 cannot explain why France had a revolution. The past history of the debt burden itself cannot either.”
  • The mechanism: Thomas J. Sargent and François R. Velde, “Macroeconomic Features of the French Revolution,” Journal of Political Economy 103(3), 1995, p. 488 (“the ability to adjust taxes did not match the king’s plans to service his debts”; constraints that “impinged particularly when it became desirable to raise taxes to service debts”), and pp. 484–485 on the incidence of privilege. Eugene N. White, Rutgers working paper 1999-04, p. 32: “Privilege shaped French society and privilege shaped the fiscal system, with vested interests it created blocking reform efforts.” The exemption was partial and negotiated, not a clean zero: White, p. 8 n. 5, and p. 16 on the vingtième.
  • The subsistence side: George Rudé, The Crowd in the French Revolution (Clarendon Press, 1959), pp. 25, 33, 67–68, and Appendix VII, whose footnote records that nearly all the wage data are for 1790.
  • Confidence is moderate under the rubric, which scores the weaker of evidence directness and construct match. Evidence directness is uneven but adequate for the ruling being made: falsifying “inevitably” requires one counter-case, and Weir’s table supplies France and Britain in the same year, in the same columns, from one reconstructor. The binding, weaker leg is construct match. “Inequality,” as the claim uses the word, is a gap between rich and poor; what the sources here measure is something else, the incidence of tax privilege and the share of a labourer’s earnings going to bread, and no figure in this chapter measures the distribution of French wealth at all. The positive half of the ruling is weaker still: the conditions listed here are drawn from a single episode, reconstructed from surviving fragments by historians who disagree with each other, and one case can show that a mechanism operated once. It cannot establish how often it operates, or what else would have to be true for it to operate again.

What the ledger will not tell you

The answer to the question this chapter opened with is that nobody knows when inequality breaks a regime, and the case everyone reaches for does not contain the answer they reach for it to supply. It contains something else. It contains a state that had made itself unable to collect from the people it protected, and had kept its books where nobody could check them, and met a winter in which bread came to take four fifths of a labourer’s effective earnings, and had to summon its own opposition into a single room in order to ask for money. Every one of those is a condition, and every one of them is a thing a country either has or does not have, and none of them is a number on a wealth distribution.

Which leaves the reader with the harder half of the work, and this chapter is not going to do it for them. The conditions are on the record. Whether they are present anywhere now, and what would count as evidence either way, is not a question a chapter about 1789 can answer, and any chapter that offered to would be selling something. What can be said is that the ledger case, examined closely, is not the case the slogan thinks it is: the crown did not fall because its nobles were rich. The crown that fell was one that could not make its richest subjects pay, and had been built so that it could not.

That is a conclusion assembled from damaged evidence by historians who contradict each other, about one country, in one year, more than two centuries ago. It is worth asking what a claim like that is really worth, and how anyone would know: whether the discipline that produced these figures can be trusted, where its knowledge is solid and where it is thin, and whether the experts are any use. That is the last question in this book. Before it, one more pathology, the markets that sell a projected future income and the dispute over who must produce the evidence behind it.