478 JOURNAL OF POLITICAL ECONOMY 90 80 France 70 60 50 40 30 20 10 1700 1720 1740 1760 1780 FIc.1.-Ratio of debt service to taxes,Britain and France,1688-1788.Sources:for Britain:Mitchell(1988);for France:Weir (1989)and the references listed in n.8. The British Experience The British pattern for this ratio was to increase during wars,when debt was incurred to pay for large military expenditures.During a war,taxes were raised to assure adequate funds to service the loans. After a war,the floating debt5 was consolidated into perpetual annu- ities,and taxes were further increased to generate a sufficient net-of- interest surplus to service the debt. Figure 2 shows the components of Britain's budget constraint dur- ing the same period.Total fiscal revenues are set against total spend- ing decomposed into military,civil,and debt service spending.The net-of-interest civil government spending is roughly constant.Mili- tary spending surges during wars.In contrast to the volatility of total government expenditures,revenues are smooth.The British govern- ment incurred large deficits in wartime and generated small but suf- ficient surpluses in peacetime.We observe a cycle of debt service rising during each war and then slowly declining with the onset of peace.Britain did not default on its debt during the 100 years follow- ing the Glorious Revolution of 1688,which reflected the existence of 5By floating debt we mean the following:in Britain,the unfunded debt(notes issued by the Exchequer or various departments,without an act of Parliament);in France, the anticipations,notes issued by the departments or financial officers on behalf of the government,typically maturing within 2 years
478 JOURNAL OF POLITICAL ECONOMY 90 80 -France 70 60- 50 4o 40 30 20 1700 1720 1740 1760 1780 FIG. 1.-Ratio of debt service to taxes, Britain and France, 1688-1788. Sources: for Britain: Mitchell (1988); for France: Weir (1989) and the references listed in n. 8. The British Experience The British pattern for this ratio was to increase during wars, when debt was incurred to pay for large military expenditures. During a war, taxes were raised to assure adequate funds to service the loans. After a war, the floating debt5 was consolidated into perpetual annuities, and taxes were further increased to generate a sufficient net-ofinterest surplus to service the debt. Figure 2 shows the components of Britain's budget constraint during the same period. Total fiscal revenues are set against total spending decomposed into military, civil, and debt service spending. The net-of-interest civil government spending is roughly constant. Military spending surges during wars. In contrast to the volatility of total government expenditures, revenues are smooth. The British government incurred large deficits in wartime and generated small but sufficient surpluses in peacetime. We observe a cycle of debt service rising during each war and then slowly declining with the onset of peace. Britain did not default on its debt during the 100 years following the Glorious Revolution of 1688, which reflected the existence of 5By floating debt we mean the following: in Britain, the unfunded debt (notes issued by the Exchequer or various departments, without an act of Parliament); in France, the anticipations, notes issued by the departments or financial officers on behalf of the government, typically maturing within 2 years
FRENCH REVOLUTION 479 30 25 20 spunod total govt spending 5 15 10 revenues civil plus debt service civil 八 1690 1700171017201730174017501760177017801790 FIc.2.-Revenues and spending in Britain,1689-1790.Total spending is decom- posed into three components:civil,debt service,and military expenditures.The three lines recorded for expenditures pertain to civil expenditures,civil plus debt service, and then total expenditures,so that the vertical distances between these lines represent, respectively,civil expenses,debt service,and military expenditures.Total revenues are depicted with small circles.Source:Mitchell (1988). mechanisms intended to make the state creditworthy (see North and Weingast [1989]for a modern account).The British king retained executive power,but the Parliament gained the powers to examine and censor the budget and to vote taxes.By 1715,the system had been refined into a method of funding by which each loan was accom- panied by a parliamentary vote for a specific tax to service the loan. Established in 1694,the Bank of England became an important element of a mechanism committing the government to pay its debts. By the mid 1720s,after the South Sea Bubble,the Bank had acquired virtual monopolies of servicing government debt and issuing notes. The Bank was designed to prevent the government from playing one lender against another.5 Its charter made it more difficult for the government to default,and the prominence of principal owners of the Bank ensured that any attempt to default would be well publi- 6 For interpretations of the Bank of England as a commitment mechanism,see Ma- caulay (1831,vol.3)and Hicks (1969,pp.93-95).See Greif,Milgrom,and Weingast (1994)for a related analysis
FRENCH REVOLUTION 479 30 25 - 20- ~0 ?L ttotal govt spending '60515- 0I 10 / reenues / 5 /civil plus debt service _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ ~~~civil 1690 1700 1710 1720 1730 1740 1750 1760 1770 1780 1790 FIG. 2.-Revenues and spending in Britain, 1689-1790. Total spending is decomposed into three components: civil, debt service, and military expenditures. The three lines recorded for expenditures pertain to civil expenditures, civil plus debt service, and then total expenditures, so that the vertical distances between these lines represent, respectively, civil expenses, debt service, and military expenditures. Total revenues are depicted with small circles. Source: Mitchell (1988). mechanisms intended to make the state creditworthy (see North and Weingast [1989] for a modern account). The British king retained executive power, but the Parliament gained the powers to examine and censor the budget and to vote taxes. By 1715, the system had been refined into a method of funding by which each loan was accompanied by a parliamentary vote for a specific tax to service the loan. Established in 1694, the Bank of England became an important element of a mechanism committing the government to pay its debts. By the mid 1 720s, after the South Sea Bubble, the Bank had acquired virtual monopolies of servicing government debt and issuing notes. The Bank was designed to prevent the government from playing one lender against another.6 Its charter made it more difficult for the government to default, and the prominence of principal owners of the Bank ensured that any attempt to default would be well publi- 6 For interpretations of the Bank of England as a commitment mechanism, see Macaulay (1831, vol. 3) and Hicks (1969, pp. 93-95). See Greif, Milgrom, and Weingast (1994) for a related analysis
480 JOURNAL OF POLITICAL ECONOMY cized.In exchange for abstaining from opportunistic behavior,the government acquired credit.? Across the Channel For France,the three sharp falls in the ratio in figure 1 each corre- spond to an episode of reimbursement suspension and default in the form of cuts in interest payments.The Spanish Succession War in 1713 marks the first episode.By 1715,the debt service had been reduced through defaults on significant parts of the floating debt. The Regency (1715-23)witnessed the "system"of John Law,a vast operation that first reimbursed the debt with bank notes that deval- ued quickly.Next,the debt was reconverted into perpetuals and life annuities,with sharply reduced capital value.Finally,because these measures proved insufficient,an interest rate cut was imposed in 1726,by which time the debt service ratio stabilized at 30 percent. The second episode occurred during the Seven Years'War in 1759, when the government converted the floating debt into perpetuals and halted scheduled reimbursements of fixed-term loans.These reduc- tions continued until the Peace of Paris forced the government to resume its obligations for a short time.Lack of funds soon curtailed compliance.Meanwhile,the floating debt had again bloated to un- manageable proportions,so a third episode started in 1770.The min- istry of Terray suspended reimbursements,converted the floating debt to perpetuals,and imposed coupon reductions on bonds of be- tween 7 percent and 50 percent.Taxes were also increased in 1771-73.When Louis XVI succeeded his grandfather in 1774,the floating debt was negligible,debt service stood at less than 40 percent of revenues,and the budget was nearly balanced:375 million livres in revenues offset 415 millions in expenditures,of which 40 millions were reimbursements of outstanding debt. These recurrent French defaults reveal different patterns of gov- ernment revenues and spending between France and England.For this period,budget data for France are not as available as for Britain. To provide a counterpart to figure 2,we constructed estimates of revenues and the components of spending for the period 1759-88 (see fig.3).8 Revenues increased sharply between 1770 and 1773,and 7 In 1797,the British government was to discover another major advantage to the institution,namely the suspension of convertibility of notes,which allowed the financ- ing of the French wars.Significantly,the government did not default on the notes and resumed the gold standard at par in 1819,in effect redeeming the notes at face value and giving a handsome return to those who had accepted the notes during the Bank restriction. 8 France did not publish the government's accounts before the nineteenth century, and the financial archives burned in 1871.The remaining information is scarce,sparse
480 JOURNAL OF POLITICAL ECONOMY cized. In exchange for abstaining from opportunistic behavior, the government acquired credit.7 Across the Channel For France, the three sharp falls in the ratio in figure 1 each correspond to an episode of reimbursement suspension and default in the form of cuts in interest payments. The Spanish Succession War in 1713 marks the first episode. By 1715, the debt service had been reduced through defaults on significant parts of the floating debt. The Regency (1715-23) witnessed the "system" of John Law, a vast operation that first reimbursed the debt with bank notes that devalued quickly. Next, the debt was reconverted into perpetuals and life annuities, with sharply reduced capital value. Finally, because these measures proved insufficient, an interest rate cut was imposed in 1726, by which time the debt service ratio stabilized at 30 percent. The second episode occurred during the Seven Years' War in 1759, when the government converted the floating debt into perpetuals and halted scheduled reimbursements of fixed-term loans. These reductions continued until the Peace of Paris forced the government to resume its obligations for a short time. Lack of funds soon curtailed compliance. Meanwhile, the floating debt had again bloated to unmanageable proportions, so a third episode started in 1770. The ministry of Terray suspended reimbursements, converted the floating debt to perpetuals, and imposed coupon reductions on bonds of between 7 percent and 50 percent. Taxes were also increased in 1771-73. When Louis XVI succeeded his grandfather in 1774, the floating debt was negligible, debt service stood at less than 40 percent of revenues, and the budget was nearly balanced: 375 million livres in revenues offset 415 millions in expenditures, of which 40 millions were reimbursements of outstanding debt. These recurrent French defaults reveal different patterns of government revenues and spending between France and England. For this period, budget data for France are not as available as for Britain. To provide a counterpart to figure 2, we constructed estimates of revenues and the components of spending for the period 1759-88 (see fig. 3).8 Revenues increased sharply between 1770 and 1773, and 7 In 1797, the British government was to discover another major advantage to the institution, namely the suspension of convertibility of notes, which allowed the financing of the French wars. Significantly, the government did not default on the notes and resumed the gold standard at par in 1819, in effect redeeming the notes at face value and giving a handsome return to those who had accepted the notes during the Bank restriction. 8 France did not publish the government's accounts before the nineteenth century, and the financial archives burned in 1871. The remaining information is scarce, sparse
FRENCH REVOLUTION 481 700 total spending 600 500% revenues 3005 米 civil plus debt service 200F 米 米米 黑黑·米 100F debt service 0 1760 1765 1770 1775 1780 1785 FIG.3.-Revenues and spending in France,1759-88.Revenues are depicted with small circles.Expenditures are decomposed as in fig.2,with successive strings of points denoting civil expenditures,civil plus debt service,and total spending,respectively.So the vertical distances between successive strings of points for expenditures represent civil,debt service,and military expenditures.Source:see n.8. grew steadily from 1776 on.The ends of the Seven Years'War and the American War for Independence triggered reductions in spending. The contrast with figure 2 is instructive.In 1763 and 1783,there was no equivalent to the British debt funding,namely a tax increase sufficient to fund the interest on the debt accumulated during the previous war.In the 1760s,tax revenues remained constant.In the 1780s,they grew too slowly,causing debt service to increase.By 1788, as in 1770 and during the Regency,the inexorable compounding of interest brought France to a fiscal crisis.9 Figure 3 shows France in the grips of some "unpleasant arithmetic"(see Sargent and Wallace 1981). and contradictory.We have constructed the estimates shown here using various pub- lished and archival sources,including the following:those published are Necker(1820), Mathon de la Cour (1788),Marion(1914-21),Clamageran (1876),Compte rendu (1788), and Compte general(1789);unpublished sources are Paris,Bibliotheque Nationale,mss. Joly de Fleury 1437 and 1442,MFF 7749,MFF 4580;and Archives Nationales,F4 1082;see also Riley (1987). 9 Over the whole period,France was at war only one year out of three.The gross-of- interest budget was in balance only briefly in the 1770s
FRENCH REVOLUTION 481 700 6001 -total spending 500 >400 O )K revenues )K- -~3000 * *| * >3 * 3 , ~~~~~~civil plus debt service *f 3 200 * i 0 ) K O 3 K )K' 100( )K 100 debt service 1760 1765 1770 1775 1780 1785 FIG. 3.-Revenues and spending in France, 1759-88. Revenues are depicted with small circles. Expenditures are decomposed as in fig. 2, with successive strings of points denoting civil expenditures, civil plus debt service, and total spending, respectively. So the vertical distances between successive strings of points for expenditures represent civil, debt service, and military expenditures. Source: see n. 8. grew steadily from 1776 on. The ends of the Seven Years' War and the American War for Independence triggered reductions in spending. The contrast with figure 2 is instructive. In 1763 and 1783, there was no equivalent to the British debt funding, namely a tax increase sufficient to fund the interest on the debt accumulated during the previous war. In the 1760s, tax revenues remained constant. In the 1780s, they grew too slowly, causing debt service to increase. By 1788, as in 1770 and during the Regency, the inexorable compounding of interest brought France to a fiscal crisis.9 Figure 3 shows France in the grips of some "unpleasant arithmetic" (see Sargent and Wallace 1981). and contradictory. We have constructed the estimates shown here using various published and archival sources, including the following: those published are Necker (1820), Mathon de la Cour (1788), Marion (1914-21), Clamageran (1876), Compte rendu (1788), and Compte ggngral (1789); unpublished sources are Paris, Bibliotheque Nationale, mss. Joly de Fleury 1437 and 1442, MFF 7749, MFF 4580; and Archives Nationales, F4 1082; see also Riley (1987). 9 Over the whole period, France was at war only one year out of three. The gross-ofinterest budget was in balance only briefly in the 1770s
482 JOURNAL OF POLITICAL ECONOMY French Fiscal Backwardness or Optimality? In figure 2,Britain looks like a simulation from a Barro (1979)tax smoothing;in figure 3,France does not.It is tempting to compare France's financial arrangements unfavorably to Britain's,but theories of dynamic Ramsey taxation instruct us to be cautious about con- demning France's recurrent defaults or praising Britain's abstinence. These theories have governments offering their creditors state- contingent,after-tax returns that respond to news about the govern- ment's prospective net-of-interest fiscal surplus.Realization of a posi- tive shock to government expenditures or a negative shock to "exogenous"government revenues results in a lower payoff and good news in a higher payoff.10 We have not attempted to match these theories to our eighteenth-century observations on French and Brit- ish finance.We would have to struggle to reconcile the timing of the French defaults with these theories,in which low after-tax returns are paid on government debt at the starts of"wars."Perhaps it could be argued that the French defaults occurred in response to reckon- ings that the prospective revenues expected to accrue with French war victory had evaporated with defeat.It would require much more work to coax from these theories an understanding of why France refused to default in 1789,despite its earlier intermittent defaults. In 1789,modernizing elements in France did not regard past and prospective government defaults as part of an optimal fiscal arrange- ment.They hoped to reform fiscal institutions to rid France of those defaults,and this is one of the reasons they welcomed the king's call to the Estates General. Snapshot of the Old Regime Laws France had the appearance of an absolute monarchy.The king cre- ated law by edicts and took executive actions by Arrets du Conseil, although not all laws emanated from the king.There were also legal traditions,with forms of common law and written law varying from province to province that preexisted royal edicts.The king con- fronted the power of a dozen Parlements,with the Paris court preem- inent,which verified that new edicts were consistent with the existing 10 This is the message of Lucas and Stokey's(1983)examples 7 and 8 and Chari, Christiano,and Kehoe's(1994)fig.6. Although Britain did not default,it did issue securities callable at the government's option.Such securities expand the range of contingencies against which a government can insure and tax smooth.Equilibrium tax-smoothing models with a government constrained to issue only callable securities have not been worked out,so we do not know how closely government callable debt would approximate a setting with a com- plete set of state-contingent securities.See also Grossman and Van Huyck(1988)
482 JOURNAL OF POLITICAL ECONOMY French Fiscal Backwardness or Optimality? In figure 2, Britain looks like a simulation from a Barro (1979) tax smoothing; in figure 3, France does not. It is tempting to compare France's financial arrangements unfavorably to Britain's, but theories of dynamic Ramsey taxation instruct us to be cautious about condemning France's recurrent defaults or praising Britain's abstinence. These theories have governments offering their creditors statecontingent, after-tax returns that respond to news about the government's prospective net-of-interest fiscal surplus. Realization of a positive shock to government expenditures or a negative shock to ''exogenous" government revenues results in a lower payoff and good news in a higher payoff.10 We have not attempted to match these theories to our eighteenth-century observations on French and British finance. We would have to struggle to reconcile the timing of the French defaults with these theories, in which low after-tax returns are paid on government debt at the starts of "wars." Perhaps it could be argued that the French defaults occurred in response to reckonings that the prospective revenues expected to accrue with French war victory had evaporated with defeat. It would require much more work to coax from these theories an understanding of why France refused to default in 1789, despite its earlier intermittent defaults." In 1789, modernizing elements in France did not regard past and prospective government defaults as part of an optimal fiscal arrangement. They hoped to reform fiscal institutions to rid France of those defaults, and this is one of the reasons they welcomed the king's call to the Estates General. Snapshot of the Old Regime Laws France had the appearance of an absolute monarchy. The king created law by edicts and took executive actions by Arrnts du Conseil, although not all laws emanated from the king. There were also legal traditions, with forms of common law and written law varying from province to province that preexisted royal edicts. The king confronted the power of a dozen Parlements, with the Paris court preeminent, which verified that new edicts were consistent with the existing '0 This is the message of Lucas and Stokey's (1983) examples 7 and 8 and Chari, Christiano, and Kehoe's (1994) fig. 6. '" Although Britain did not default, it did issue securities callable at the government's option. Such securities expand the range of contingencies against which a government can insure and tax smooth. Equilibrium tax-smoothing models with a government constrained to issue only callable securities have not been worked out, so we do not know how closely government callable debt would approximate a setting with a complete set of state-contingent securities. See also Grossman and Van Huyck (1988)