¹ Whether borrowing by Britain to finance war efforts in the late eighteenth and early nineteenth centuries affected interest rates and private investment is by no means a recent issue. Mill is among the nineteenth-century writers who suggested that war debts could, under certain circumstances, displace private capital accumulation. See Mill, John Stuart, Principles of Political Economy (1862; New York, 1909 printing of 5th edn.), pp. 873–75. The recent interest in historical crowding out is related to U.S. experience with high budget deficits in the 1980s. A number of studies have examined the potential for deficits, or government expenditure, to crowd out private investment.Google Scholar See, for example: Evans, Paul, “Do Deficits Raise Interest Rates?” Journal of Monetary Economics, 20 (09 1987), pp. 281–300CrossRefGoogle Scholar; Dwyer, Gerald P. Jr, “Federal Deficits, Interest Rates and Monetary Policy,” Journal of Money, Credit, and Banking, 17 (11 1985, Part 2), pp. 655–81CrossRefGoogle Scholar; Barro, Robert J., “Government Spending, Interest Rates, Prices, and Budget Deficits in the United Kingdom, 1701–1918,” Journal of Monetary Economics, 20 (09 1987), pp. 221–47CrossRefGoogle Scholar; Tatom, John A., “Two Views of the Effects of Government Budget Deficits in the 1980's,” Federal Reserve Bank of St. Louis Review, 67 (10 1985), pp. 5–16Google Scholar; and Bradley, Michael D., “Government Spending or Deficit Financing: Which Causes Crowding Out?” Journal of Economics and Business, 38 (08 1986), pp. 203–14. Bradley includes a brief summary of the results from 14 earlier studies of the link between deficits and interest rates.CrossRefGoogle Scholar

² Williamson, Jeffrey G., “Why Was British Growth So Slow During the Industrial Revolution?” this JOURNAL, 44 (09 1984), pp. 687–712Google Scholar, see especially p. 712. It is wartime borrowing that is at the heart of the argument. For a study of crowding out during the Civil War, World War I, and World War II in the United States, see Evans, Paul, “Do Large Deficits Produce High Interest Rates?” American Economic Review, 75 (03 1985), pp. 69–87.Google Scholar An interesting classical view on why the “want of parsimony” among sovereigns during peace gives rise to wartime borrowing in the first place is in Smith, Adam, The Wealth of Nations (1776; New York, 1937), pp. 861–62. Smith also noted that “in the war which began in 1688 … the foundation of the … enormous debt of Great Britain was first laid” (see pp. 873–74).Google Scholar

³ Heim, Carol E. and Mirowski, Philip, “Interest Rates and Crowding-Out During Britain's Industrial Revolution,” this JOURNAL, 47 (03 1987), pp. 117–39.Google Scholar

⁴ Ibid., p.118.

⁵ Ibid., pp.121–22.

⁶ Williamson, Jeffrey G., “Debating the Industrial Revolution,” Explorations in Economic History, 24 (07 1987), p. 288CrossRefGoogle Scholar; see also Williamson's, response to HM in “Has Crowding Out Really Been Given a Fair Test? A Comment,” this JOURNAL, 47 (03 1987), pp. 214–16. The same point regarding the inadequacy of the expectations variable was also made to the present authors by Thomas Chiang.Google Scholar

⁷ HM hold to this concept of business investment. See “Interest Rates and Crowding Out,” p. 137–38.Google Scholar

⁸ Ibid., p. 121. HM chose the rate on the popular and liquid East India Company bonds to “reflect conditions in the short-term private bond market” (p. 121). They chose the real consol yield as a measure of conditions in the long-term debt market. On this issue, see Homer, Sidney, A History of Interest Rates (2nd edn., New Brunswick, 1977), pp. 159–60. Because the choice of price index did not affect HM's results in any crucial way, we use only the Schumpeter-Gilboy consumer price index. The data are reported in “Interest Rates and Crowding Out,” table 1 (p. 120) and table 3 (p. 124).Google Scholar To accommodate long lags in estimating inflation expectations, we took additional Schumpeter-Gilboy price data from Mitchell, B. R. and Deane, P., Abstract of British Historical Statistics (Cambridge, 1962), p. 469.Google Scholar

⁹ HM, “Interest Rates and Crowding Out,” p. 126, 127. fn. 16. HM also cited a potential complication in the data on real government debt due to debt-conversion operations beginning in 1808 (p. 126). The real net receipts from borrowing series corrects for these defects, limiting results to the years 1782 to 1816 (see also p. 123). Since the interpretation of HM's empirical results depends on the assumption of simultaneity bias, we tested for its presence when change in real debt is the debt variable, using the procedure outlined in Ramsey, James B. and Schmidt, Peter, “Some Further Results on the Use of OLS and BLUS Residuals in Specification Error Tests,” Journal of the American Statistical Association, 71 (1976), pp. 389–90. The results supported the hypothesis of simultaneity bias for change in real debt but not for real net receipts from borrowing.CrossRefGoogle Scholar

¹⁰ “Interest Rates and Crowding Out,” p. 121.Google Scholar

¹¹ An anonymous referee noted a difficulty in estimating equation 2 with the interaction term: the nominal yield appears on both sides of the equation. Ignoring the interaction term could bias estimates of the coefficients, particularly a₂ in equation 4. Accounting for the interaction, however, introduces its own substantial bias. Furthermore, low coefficients on inflation expectations in the nominal-rate equations reported below suggest that, if bias is present, it is not very substantial.Google Scholar

¹² While it is possible to correct for autocorrelation using one of several econometric procedures, the preferred approach to autocorrelated errors is first to review the theoretical specification. The reason is that serially correlated omitted variables can induce serial correlation in the error term, which captures the effects of omitted variables along with random shocks.Google Scholar

¹³ Humphrey, Thomas M., “The Early History of the Real/Nominal Interest Rate Relationship,” Essays on Inflation (5th edn., Richmond, 1986), p. 151.Google Scholar

¹⁴ See the discussion of Irving Fisher's Appreciation and Interest (1896) in ibid., pp. 156–57.

¹⁵ Ashton, T. S., in An Economic History of England: The 18th Century (1955; London, 1972), p. 199, notes that it was primarily agricultural prices which were volatile.Google ScholarDeane, Phyllis and Cole, W. A. show that the period of 1790 to 1825 exhibited especially severe short-run reversals in price movements; see British Economic Growth, 1688–1959 (Cambridge, 1967), p. 351, fig. 7.Google Scholar

¹⁶ The price trend was calculated using a five-year moving average, including the current rate of inflation and four lagged values.Google Scholar

¹⁷ See HM, “Interest Rates and Crowding Out,” p. 120, table 1.Google Scholar

¹⁸ An anonymous referee suggested that inflation be explained using all the available information, including information on debt and nominal yields as well as information on past inflation. We also included the growth rate of money but this did not add to the significance of the regression, perhaps because debt and money are correlated. On the role of debt as money, see the discussion below in section 3. The same referee also suggested that the data on price included a possible regime change when the British economy went from peace to war in 1793. Estimating separate equations did show some changes in coefficients but an F-test for the significance of separate regressions showed that allowing for a regime change did not significantly increase the ability to explain inflation. In testing for crowding out, a peacetime/wartime inflation-expectations variable in unreported nominal-yield regressions did not change the results for the crucial borrowing variable.Google Scholar

¹⁹ (New York, 1930), chap. 19. See especially pp. 416–25 for a study of the relation between inflation and nominal consol yields in Britain from 1820 to 1924. Fisher argued that realized, or cx post, real rates would be more volatile than nominal rates because “men are unable or unwilling to adjust at all accurately and promptly the money interest rates to changed price levels” (p. 415).Google Scholar

²⁰ Ibid., pp. 424–25.

²¹ Estimating current inflation (PP₁, rather than PP^e₁) as a 25-year distributed lag of past inflation for 1785 to 1816 will give an adjusted R² of only about 0.020. Comparing this result with equation 5 shows that the distributed lag approach is indeed inferior as an inflation forecasting equation. Even if information on government borrowing and nominal yields were not included in equation 5, its adjusted R² would only drop to 0.285, leaving it well above that for the distributed-lag equation.Google Scholar

²² HM's choice of the date on which to measure interest rates, the first Wednesday of April of each year, makes the issue of lags all the more important. Much of the government borrowing in 1782, for example, probably took place after the date in question, assuming a fairly even rate of borrowing. As a result one ought to allow not only for a contemporaneous effect but also for a one-year lagged effect of borrowing on interest rates. See HM, “Interest Rates and Crowding Out,” p. 119.Google Scholar

²³ See, for example, the studies by Evans, “Do Large Deficits Produce High Interest Rates?” and “Do Deficits Raise Interest Rates?” which also assume lagged borrowing effects.Google Scholar

²⁴ An indication that borrowing could have a lagged effect on interest rates emerges from a report on the suspension of specie payment in 1797, an event that is closely related to the heavy wartime borrowing by the British government. In an anonymous tract, “Note on the Suspension of Cash Payments, at the Bank of England, in 1797,” in McCulloch, John R., ed., A Select Collection of Scarce and Valuable Tracts and Other Publications, on Paper Currency and Banking (1857; New York, 1966), pp. 95–96, a contemporary writer highlighted the severity of the financial crisis leading to the suspension and outlined its probable causes, as well as the Bank of England's response: By far the most important crisis in the history of the paper currency of Great Britain took place in 1797. Owing partly to events connected with the war in which we were then engaged, to loans to the Emperor of Germany, to bills drawn on the treasury by the British agents abroad, and partly and chiefly, perhaps, to the large advances made by the Bank of England to Government, the exchange became unfavorable in 1795, and in that and the following year large quantities of specie were demanded from the Bank. No doubt, however, the ultimate crisis was wholly owing to political causes. Alarms with respect to invasion, and reports of descents, said to have been made on the coast, became exceedingly prevalent in the latter part of 1796, and the beginning of 1797. This produced a strong desire among many individuals, but chiefly among the small farmers and retail dealers, to convert as much as possible of their property into cash…Demands for supplies of cash poured in upon the Bank of England from all parts of the country; and the stock of coin and bullion in her coffers, which amounted to £7,940,000 in March, 1795, was reduced on Saturday, the 25th of 02 1797, to £1,272,000 with every prospect of a violent run taking place on the following Monday. In this emergency, a meeting of the Privy Council was held, when it was resolved to suspend payments in cash at the Bank until the sense of Parliament could be taken on the matter. The important point to observe from this account is an implied lag in the effect of governmental borrowing. Borrowing for the war with France, which began in 1793 and became quite heavy by 1795, was one underlying cause of the financial crisis in late 1796 and the subsequent suspension of specie payment early in 1797. The clear suggestion here is that the effect of wartime borrowing may have operated on the financial community, in particular on the nominal rate of interest, with a lag.Google Scholar

²⁵ An anonymous referee suggested that a long delay of two years in the effect of borrowing on rates “is inconsistent with the simplest principles of asset pricing [which] require that predictable changes in asset prices be smooth. Large predictable changes in interest rates due to information about government activity last year would imply forecastable profits.” To avoid this difficulty, estimates of the yield equation reported below were first run with a smooth lag distribution running from zero to three years. Nevertheless, the coefficient on borrowing at a two-year lag was, for the nominal India bond yields, the first consistently significant one. In estimates for nominal consol yields, however, the one- and two-year lags were both significant; this finding would be consistent with asset pricing theory, given the choice of date for measuring the interest rate (see the discussion of this point earlier in this section). More evidence on the institutional aspects of financial markets is needed to reconcile the empirical data with a reasonable theoretical proposition.Google Scholar

²⁶ The rule for respecification was to keep a variable if its coefficient had a t-statistic greater than one. This is approximately the rule used in several recently developed statistical criteria for model selection. See Judge, George G. et al. , The Theory and Practice of Econometrics (New York, 1980), pp. 420–22. Given the available data on real net receipts from borrowing (1782 to 1816), a lag of three years on government borrowing left 1785 as the starting year for the sample.Google Scholar

²⁷ The results do not differ when change in real debt is substituted for real net receipts, the variable preferred by HM.Google Scholar

²⁸ The results for lagged borrowing held up in tests with alternative expectations assumptions: an autoregressive expectations variable which only considered two lagged values of inflation; a perfect-foresight approach; and Fisher's distributed-lag approach, using polynomial lags of 5, 10, 15, 20, and 30 years.Google Scholar

²⁹ Ricardo, David, Principles of Political Economy and Taxation (3rd edn., 1821; reprinted London, 1908), p. 281.Google Scholar

³⁰ “Debating the Industrial Revolution,” p. 288, and “Has Crowding Out,” pp. 214–16.Google Scholar

³¹ HM, “Interest Rates and Crowding Out,” p. 127.Google Scholar

³² Evans, “Do Large Deficits Produce High Interest Rates?” pp. 72–73, for example, found no empirical connection between government deficits and interest rates for the United States during the Civil War period. His explanation of this result lists arguments which are similar to some of those to be discussed below.Google Scholar

³³ See Ricardo, David, “The High Price of Bullion a Proof of the Depreciation of Bank Notes,” The Works and Correspondence of David Ricardo, Sraffa, Piero and Dobb, M. H., eds. (Cambridge, 1962), vol. 3, pp. 47–98Google Scholar; see also Blake, William, “Observations on the Principles which Regulate the Course of Exchange: and on the Present Depreciated State of the Currency” (1810), in McCulloch, , ed., A Select Collection of Scarce and Valuable Tracts and Other Publications, on Paper Currency and Banking, pp. 523–24.Google Scholar

³⁴ An Economic History of England, pp. 177–78.Google Scholar

³⁵ This could be illustrated by a rightward shift of both the IS and the LM curves so as to leave the nominal interest rate unchanged. Until inflation expectations began to catch up, the nominal rate would not adjust.Google Scholar

³⁶ See Ricardo, , “Funding System,” in Works and Correspondence, vol. 4, pp. 185–88.Google ScholarBarro, Robert J., in “Are Government Bonds Net Wealth?” Journal of Political Economy, 82 (11/12 1974), pp. 1095–117, labeled this the Ricardian equivalence principle, whereby increased taxes or debt to finance increased government expenditure would have the same effects on the economy.CrossRefGoogle Scholar For a helpful account of this and other issues regarding classical thought on government debt, see Rowley, C. K., “Classical Political Economy and the Debt Issue,” in Buchanan, J. M., Rowley, C. K., and Tollison, R. D., eds., Deficits (Oxford, 1986), pp. 49–74, esp. 65–69.Google Scholar For a critical discussion of Ricardian equivalence, see Tobin, James, Asset Accumulation and Economic Activity:Reflections on Contemporary Macroeconomic Theory (Chicago, 1980), chap. 3.Google Scholar

³⁷ Williamson, “Why Was British Growth So Slow,” pp. 100–1.Google Scholar

³⁸ Lipson, E., The Growth of English Society (New York, 1950), p. 357. Clapham is more conservative; he puts the “recognized membership” in the “year of waterloo” at nearly a million but notes that this did not include members of informal secret orders.Google Scholar See Clapham, J. H., An Economic History of Modern Britain (Cambridge, 1930), p. 211.Google Scholar

³⁹ Mantoux, Paul, The Industrial Revolution of the Eighteenth Century (revised edn., New York, 1961), p. 450.Google Scholar

⁴⁰ Cotton production, however, was in the midst of a period of great innovation, and the rate of growth of output increased. See Murphy, Brian, A Hictory of the British Economy: 1088–1970 (London, 1973), p. 416, table 11.5.Google Scholar

⁴¹ Ibid. Murphy's evidence on pig-iron production supports HM's (“Interest Rates and Crowding Out,” pp. 136–37) contention that the mix of output may have been changed by wartime expenditures. For a contrary view, see Williamson, “Why Was British Growth So Slow,” p. 188.Google Scholar

⁴² Evans, Paul, “Are Consumers Ricardian? Evidence for the United States,” Journal of Political Economy, 96 (10 1988), pp. 983–1004.CrossRefGoogle Scholar

⁴³ Crafts, N. F. R., “British Economic Growth, 1700–1850: Some Difficulties of Interpretation,” Explorations in Economic History, 24 (07 1987), p. 247.CrossRefGoogle Scholar

⁴⁴ An Economic History of England, p. 193.Google Scholar

⁴⁵ “Why Was British Growth So Slow,” p. 290.Google ScholarHM, “Interest Rates and Crowding Out,” pp. 133–34, also noted the role of the Dutch and other Europeans in lending to Great Britain, but they cited this as evidence against the neoclassical model rather than consistent with it. For a further discussion of financial integrationGoogle Scholar, see Neal, Larry, “Integration of International Capital Markets: Quantitative Evidence from the Eighteenth to Twentieth Centuries,” this JOURNAL, 45 (06 1985), pp. 219–26.Google Scholar

⁴⁶ An Economic History of England, p. 27.Google Scholar

⁴⁷ Ricardo, Principles of Political Economy, p. 281. See the quote in the conclusions to section I.Google Scholar

⁴⁸ McCulloch, J. R., Principles of Political Economy (2nd edn., Edinburgh, 1843), p. 520.Google ScholarMacleod, Henry Dunning, History of Economics (London, 1896), p. 470, gives further evidence of the ceiling leading to high rates on various occasions: “The experience of several commercial crises had demonstrated that in consequence of the law attempting to prevent persons paying more than 5 per cent. for a loan of money, they often had to pay 50, 60, and 70 percent by the methods they were forced to adopt,” No doubt such high rates had a chilling effect on private borrowing.Google Scholar

⁴⁹ Mill, Principles, p. 874.Google Scholar

⁵⁰ “Interest Rates and Crowding Out,” pp. 138–39.Google Scholar

⁵¹ An Economic History of England, pp. 28–29.Google Scholar Ashton cites this story from Fitton, R. S. and Wadsworth, A. P., The Struts and the Arkwrights, 1758–1830 (New York, 1958), p. 30.Google Scholar

⁵² HM, “Interest Rates and Crowding Out,” p. 117. Their point, though, was that the evidence did not support this view.Google Scholar

⁵³ For the contrary view, see ibid., p. 129.