5. Taking Stock: Monetary and Banking Policy in the Second Reign

Friedman and Schwartz (1963).

Notable examples are Cavalcanti (1890, 1893), Wileman (1896), Calógeras (1910), Ortigão (1914), Neuhaus (1975) and, more recently, Franco (2017).

Peláez and Suzigan (1976).

This appears to have been the case in 1853, when the money supply (M1), measured by the sum of notes held by the public and demand deposits, was in the order of 47,000 contos and ‘close to 30,000 contos in specie was to be found in the Brazilian market’ (Calógeras 1910, p. 110). The second half of the 1850s witnessed the largest amount of specie struck by the Brazilian mint, with 21,000 contos in gold coins and 5800 contos in silver coins put into circulation between 1854 and 1859. This comprised approximately 30% of the average outstanding circulation of notes (issued by both the Treasury and private banks) at the time, and an equal share of the average money supply as measured by notes held by the public and demand deposits. Total coinage of gold pieces in the Brazilian mint between 1850 and 1888 amounted to 45.7 thousand contos, while another 20.9 thousand contos in silver coins were put into circulation in the same period. Copper and silver coins, used as small change and issued from 1868 and 1872 onward, respectively, totaled 7.6 thousand contos. Figures in Cavalcanti (1893, vol. 2, p. 327). Unfortunately, an annual series for the stock of specie in circulation (as opposed to specie coined at the mint) is not available. Throughout the period under study here, specie was often driven out of circulation, being either exported or hoarded. This precludes any attempt to arrive at a more precise estimate of the monetary base that would result from adding figures for specie in circulation (including British sovereigns and other foreign coins, which were legal tender) to the published series for notes issued by the government and issuing banks. For a discussion, Abreu and Lago (2001, pp. 363–71).

In connection to this, Schulz mentions yet another possible correction to the Peláez and Suzigan estimates of the money supply, but relating to the final year of the Empire (1889). According to the author, the exchange-rate parity then achieved provided for the entry in circulation of the equivalent of £ 9 million in specie, or one-third of the money supply at the time. He goes on to cite page 336 in Cavalvanti (1890) as his source, although the latter does not provide this figure but, solely, estimates of gold coins minted in Brazil in different points in time—not specie in circulation. Schulz (2013, p. 164). Still, fragmentary evidence suggests that there was, indeed, an increase in the volume of metal coins in circulation (and, therefore, of the money supply) in Brazil in 1888 and 1889. Over these two years, there was a net inflow of £ 2.9 million in silver and gold coins from Britain, part of which was minted into Brazilian currency and put into circulation. The Brazilian mint produced just over 4.5 thousand contos in silver coins in 1888 and 1889, valued at £ 487,000. There are no detailed data on gold coins minted in Brazil in those two years, although Cavalcanti (1893, vol. 2, p. 315) provides a figure for 1.9 thousand contos (equivalent to £ 205,600) in gold pieces coined in 1888–9. Data on specie imported from Britain in Lago (1982). Coins minted in Brazil in Amato et al. (2008).

The Brazilian population roughly doubled between 1850 and 1889, from an estimated 7.3 million to 14.3 million according to the 1890 census. Brasil, IBGE (1990, pp. 33–4).

For the 1870–1913 period, Catão (1992b) provides a carefully constructed wholesale price index.

As discussed in the text, this is, admittedly, a restrictive definition of the monetary base, as it leaves out metallic currency in circulation and vales (promissory notes) issued by a number of agents at the time, including private banks, companies and, even, individuals.

At its peak, in December 1861, notes issued by the banks created by government decree in 1857–8 reached 13.5 thousand contos, or 16.5% of total notes in circulation. Thereafter—and because of their incapacity to comply with the rules set out by the 1860 Law of Impediments—, their combined issues declined significantly and would constitute but a marginal share of total notes in circulation (see Table A1 in the Appendix).

Chapters 2, 3 and 4 present this history in detail.

For details, see Chap. 4 in this book and Summerhill (2015, ch. 4).

Figures on funded domestic debt outstanding in Levy (1995). Data originally published in milréis and converted to pounds sterling at the average monthly rate of exchange at each point in time over the whole series.

Figures for apólices in circulation refer to March 1866 and March 1889. Levy (1995, pp. 243–9).

This figure does not include the 1889 conversion loan.

For Mauá’s views on monetary matters, Fernandes (1974). Given the weight of receipts from coffee in total exports under the Empire (roughly 50%), much attention has been given to studying its influence in determining the exchange rate in the period. In this regard, Furtado’s classic contribution, by pointing out the instability in the international demand for exports from primary-producing countries, highlights the serious consequences that sudden shifts in demand would have on export earnings from the periphery and, hence, the exchange rate of those countries. Furtado (1970, pp. 164 ff.). Furtado actually reinforces an argument made earlier by Wileman, relating oscillations in the value of the milréis and the price of coffee. Wileman (1896). Later, this relationship was challenged by Delfim Netto (1979, p. 26), who concluded that ‘in the long run the exchange rate was relatively little affected by the price cycles (of coffee)’. Since the 1980s, there were renewed attempts to analyze the main determinants of the exchange rate in imperial Brazil. Most of these works sought to test the extent to which the foreign value of the milréis followed changes in coffee prices, monetary policy, capital flows, the terms of trade or a combination of these variables. The results are inconclusive, but most of the contributions tend to indicate that both domestic influences—linked to coffee output, money supply, and wage setting behavior—as well as international ones, via foreign lending and the international demand for Brazil’s exports, had a bearing on the country’s exchange rate. See Cardoso (1981), Franco (1986), Fishlow (1987), and Catão (1992a, chapter VI). On the difficulties in estimating the main determinants of the exchange rate in late nineteenth-century Brazil, Franco (1991, pp. 77–80).

For a similar discussion in connection with the political economy of the exchange rate during the First Republic (1889–1930), Triner (2000).

Topik (1994).

In the case of coffee, Brazilian exports constituted a significant share of the world total and, hence, exercised considerable influence on the world price of the commodity. For a discussion, Abreu and Tamega (2005).

The above discussion of the political economy of exchange rate and monetary policy is open to a few qualifications. The first point concerns the relative sensitivity of tradable producers to exchange-rate movements, which, in turn, depends on the price elasticity of demand for their products. In the Brazilian case, the coffee sector was characterized by quasi-monopoly power, coupled with a low price elasticity of demand for its product. This feature enabled coffee producers to weather phases of exchange-rate appreciation better than other exporters. Also, foreign investors, when bringing capital to Brazil, would initially benefit from a depreciated milréis. The income stream from those very investments, on the other hand, would be greater in foreign currency terms the more appreciated the rate of exchange. Finally, in a fixed-rate regime such as the gold standard, divisions might arise within a group of exporters if there is mounting pressure for a competitive devaluation of the currency. In this case, the policy decision will involve choosing between those with long-term contracts, that is, whose sale prices in foreign currency might have been set way beforehand—and who will tend to be against the instability that devaluation introduces—and other exporters who would gain immediate competitiveness as a result of a currency devaluation. Frieden (1994, pp. 85–6).

See, for example, the debates in the 1878 agricultural congress held in Rio, reprinted as Anais do Congresso Agrícola (1988).

This is what appears to be implied by Levy and Andrade, when they contend that financing of the agricultural sector via factors (comissários) ‘reasserted the dominance of circulation over production, ensuring the control of the latter by mercantile capital’. Levy and Andrade (1993, p. 25).

Cf. Silva Netto (1865, Introduction).

Speech given by Minister of Finance Torres Homem in 1859, cited in Andrada (1923, p. 90).

Sweigart (1987).

See also Lago (2019) on the difficulty faced by creditors in foreclosing rural property in imperial Brazil.

This major obstacle in the way of the development of a long-term market for rural loans was clearly diagnosed by a contemporary. See the chapter “Do Crédito Rural e Sistema Hipotecário” (“On Rural Credit and the Mortgage System”), in Franco (1984, pp. 111–7).

On the falling prices for slaves in Brazil as contemporaries perceived the inevitability of abolition (without compensation), Mello (1978).

This “liquidity consciousness” was reinforced by the recurrent panics and crises that characterized the nineteenth century. Fears of a potential mismatch between the maturities of bank assets and liabilities has also been identified as a source of the short-termism of which nineteenth-century British banks have often been accused. Collins (1991, pp. 16–7).

It has been suggested that a further reason for the relatively high cost of credit in nineteenth-century Brazil was its underdeveloped banking system. The latter, in turn, stemmed in part from the strict control exercised by the government, including the Council of State, over the entry of new players in the market. Ultimately, political power concentrated in the hands of the Executive was used to shield from competition incumbent banks (including the Bank of Brazil), whose boards of directors often included national political officeholders. For an elaboration, Summerhill (2015, ch. 7).

The original argument is from Gerschenkron (1962, pp. 13 ff).

Indeed, this was perfectly understood by some contemporaries. Commenting on the government’s plans to allow the Banco Commercial do Rio de Janeiro and Mauá’s Bank of Brazil to increase their issue of vales, in order to alleviate the monetary pressure felt in 1853, deputy Bandeira de Mello noted that the measure was fruitless: ‘if the bills offered for discount do not have credit, if they are not sufficiently guaranteed, banks will not discount them, for banks cannot compensate for the lack of credit’. Session of 13 June, in ACD 1853, Tome II, p. 181.

The first Bank of Brazil was created in October 1808 and its operations started one year later. Throughout its twenty-year life, it was beset by demands from the State that, ultimately, led to its closure and a protracted process of liquidation, running from 1829 to 1835. For details, Franco (1984, ch. 1), Vieira (1962), Peláez (1975), and Gambi (2017).

Articles 9 and 10 of the said decree determined that ‘the establishment of any corporation with the purpose of engaging in banking operations shall only be authorized once a quarter of its shares have been subscribed; however, if this total has not been realized within the time marked by its statutes, it shall be dissolved (…)’.

Speech delivered in the Chamber of Deputies on 20 June 1853, cited in Franco and Pacheco (1979, Vol. II, p. 106).

Speech delivered in the Chamber of Deputies on 18 June 1853, cited in Franco and Pacheco (1979, Vol. II, pp. 101–2).

This was the opinion expressed by a Mr. Robert Clinton Wright at the enquiry into the 1857 commercial crisis. See Relatorio da Commissão de Inquerito de 1859 (undated, pp. 111–2).

In fact, official recognition of the role of deposits as part of the money supply would only come about when the government passed Decree 3339 (10 November 1864), which stipulated that deposits held in the Bank of Brazil in current account would henceforth count as part of its issues. The same rule would apply to three-fourths of total interest-bearing deposits. Cavalcanti (1893, Vol. 2, p. 279). In the United States, the procedure to include bank deposits with note issues when setting specie requirements was pioneered by the 1842 Louisiana banking act. Hammond (1957, p. 695).

‘In my opinion, receipts (...) issued by the bankers who failed represented an actual issue in disguise, one thousand times more dangerous and damaging to the public than those made by banks; because, although they acted as currency in commercial transactions, they had no limit whatsoever, except the individual credit of those who signed them; it was but an illicit means to circumvent the dispositions of the law of 22 August 1860’. Statement made by Sebastião Ferreira Soares for the inquiry into the 1864 crisis, in Relatorio da Comissão de Inquerito de 1864 (1865, Annex C, Part I, p. 51).

This point serves to highlight the fact that to focus exclusively on traditional monetary aggregates to establish the degree of liquidity in the economy can be misleading, as argued by Kindleberger (1996, pp. 49–51). For example, Sweigart remarks that in 1865 ‘a Rio merchant had estimated that coffee factors had lent planters as much as 90,000 contos, a sum only slightly less than the total currency held by the public over the entire country. Not all of the planter drafts ended up in banks; drafts accepted by well-known planters and factors circulated as currency throughout the coffee provinces’. Sweigart (1987, p. 135). It is difficult to assess the extent to which planter drafts behaved as money, as well as the amount of outstanding obligations of this nature in other points in time. Still, their sheer volume in 1865 should add a measure of caution to any analysis of monetary trends based exclusively on aggregates such as M1 or M2. Goldsmith (1986) puts the total amount of deposits held at private banks in 1864 at somewhere between 25–50 thousand contos (see Table 5.1). When these other components are added to the estimates of the money supply provided by M1 and M2, it becomes clear that the Brazilian economy was not suffering from a monetary crunch in the early 1860s.

On the free banking experience of Scotland, White (1984, 1991).

It has been remarked that at that time ‘only in England and Scotland were joint-stock companies generally subject to unlimited liability. In America, as well as on the Continent, the principle of limited liability became the general rule right from the beginning’. Cf. Smith (1936, p. 37, footnote 2).

Barman points out that Brazil was not alone at the time in making incorporation depend on official approval, as the same applied to France under its 1807 Commercial Code. See Barman (1981). Free incorporation was not introduced in the North German Confederation and Austria until 1870 and 1899, respectively. Cameron (1972, p. 17).

In a way, then, joint-stock issuing banks were invested with a double privilege, that of limited liability—which benefited shareholders—and that of issuing notes.

As seen in Chap. 3, the 1860 Law of Impediments imposed restrictions not only on the operation of joint-stock banks, but on corporations in general. Additionally, it strengthened the provisions of the Commercial Code, by laying down for the first time clear penalties (in the shape of fines) on those responsible for failing to obtain a charter before embarking on commercial ventures (Art. 2, para 1).

As expected, this provision could cause problems, as in the case of a respectable family from Rio, left stranded in a train station in the outskirts of the city, for want of legal tender currency to pay the fare. As it turned out, all they had in their pockets was a substantial amount of notes issued by the Banco Commercial & Agrícola, which the company refused to accept. This episode was mentioned during debates held in the Chamber of Deputies. See session of 29 May, in ACD 1858, Tome I, p. 149.

Minister of Finance Ferraz criticized the banks’ issues of notes of small denomination on the grounds that they fell into the hands of the lower classes, who used them in daily transactions. This subjected the paper to wear and tear, when not total loss, something that would ultimately benefit the issuing bank, who would be free from redeeming them. RMF 1859, p. 69.

This prompted the adoption of Article 4 of Decree 2685 (10 November 1860), which imposed on the Bank of Brazil the requirement to withdraw from circulation notes of smaller denomination, should it fail to resume (gold) convertibility within six months.

Similar problems arose in the South of the United States, where states struggled to regulate the private issue of so-called shinplasters. Schweikart (1987, p. 115).

Throughout most of the period under discussion here, import duties were not collected on an ad valorem basis. Instead, they consisted of a fixed amount (a lump sum) charged on “official values” of each merchandise as detailed in the tariff schedule. In the presence of exchange-rate swings, the market value of a given item could differ—at times, markedly—from “official values”, causing importers to face tariff rates that were quite distinct from what they anticipated (and, sometimes, from the government’s original intent). For a discussion, Villela (2005).

At the time, the degree of protection afforded to import-substituting industries by a depreciated milréis, as well as its influence on the imports of capital goods (and, hence, domestic investment), were not significant issues in the political economy of the exchange rate.

This was the third time the legal parity of the milréis was changed. Historically, the par rate of exchange had been 67.5d. The 27d/1$000 rate replaced the previous parity established in 1833 (at 43.5d/1$000) and would thereafter—and well into the republican period—constitute a veritable “Holy Grail”, a mythical rate of exchange that would be doggedly pursued by the government.

In addition, as shown in Figure 5.5, in the case of the smaller issuing banks, note issues could be partly backed by shares of companies enjoying profit (dividend) guarantees from the government.

On the British monetary controversies, Viner (1937) and Glasner (1989, 1992).

Saes (1986) and Andrade (1987).

Major politicians identified with the metalistas were Angelo Muniz da Silva Ferraz (President of the Council of Ministers and Minister of Finance from August 1859 to March 1861), Paulino José Soares de Sousa (Visconde de Uruguai) and José Maria da Silva Paranhos (Visconde de Rio Branco). With Souza Franco’s ideas sided Irineu Evangelista de Souza (Mauá), João Lustosa da Cunha Paranaguá and José Antonio Saraiva, to name the most important. Andrada (1923, p. 112).

The 1856 edition of Gilbart’s 1827 book A Practical Treatise on Banking was translated by L. J. de Oliveira e Castro as Tractado Practico de Bancos (Rio de Janeiro: Garnier, 1859, 3 vols.). I am indebted to Carlos Gabriel Guimarães for drawing my attention to the Brazilian edition of the Gilbart book.

This similarity was not lost on a contemporary author (and professor at the Recife Law School), who remarked that ‘every one of our Ministers of Finance (…), advocates of monopoly or plurality of banks, from Torres Homem to Souza Franco, all of them are blind admirers of gold’. Cf. Milet (1875, p. 25).

See, for example, Saes (1986, Part I), and Andrade (1987).

For a similar contemporary defense of plurality of issue with convertibility of (non-legal tender) banknotes, Hoyer (1880, pp. 38–40).

Andrade (1987).

An early example is Normano (1935, p. 190).

His participation in the efforts to prop up the rate of exchange in the wake of the 1857 crisis bear witness to this. For details, see Chap. 2.

Still, this did not stop him from operating in Uruguay, where notes issued by the Mauá Bank were redeemable in specie.

Mauá’s defense of inconvertible banknotes had already been made public much earlier, in his testimony to the enquiry into the 1857 commercial crisis. See Relatorio da Commissão de Inquerito de 1859 (undated, Annex A, p. 96).

See Banco do Brasil, Parecer Apresentado á Assembléa (1861).

Ibid., pp. 14–5.

Ibid., pp. 16–7. As noted by many authors, the gold standard operated differently in the countries at the center and those in the periphery of the world economy, the latter being frequently hit by convertibility crises, devaluations, and internal dislocations. In effect, borrowing countries were not as capable as core ones to control the rate of their capital imports and exports. For the periphery, buoyant economic activity led to merchandise imports picking-up and, eventually, to a trade deficit. In the early stages of the cycle, this trade deficit would be financed by (foreign) capital inflows which, by their inherent volatility, were bound to return to their countries of origin at news of panics and the subsequent increases in Bank (of England) Rate. Given the lagged response of import demand to credit conditions, this capital flight could result in severe reserve losses that, under the gold standard, would bring about an equally strong monetary contraction. Depending on the magnitudes involved, that is, the size of the gold drain relative to the domestic monetary base, the ensuing deflation could be considerable. In short, capital flows tended to be pro-cyclical, that is, they would increase in boom times and as soon as the slump set in (but merchandise imports were still on the rise), they would return to the centers from where they had originated. This would aggravate the natural instability that derived from the periphery’s frequent reliance on the exports of one or two primary products. For an elaboration of this argument, Ford (1962). For a discussion of these issues as applied to early twentieth-century Brazil, Fritsch (1988). Furtado had already made this point in his classic 1959 book, when describing the economy of nineteenth-century Brazil and its incapacity to adapt to the rules of the gold standard in the period. See Furtado (1970, ch. 27).

Soares’ reference to “exports in excess of imports” (a trade surplus), as he himself elaborates in his book, would refer in modern parlance to the need for a balance of payments surplus if gold and silver coins were to be retained domestically.

Soares’ Notas Estatísticas was originally published in 1860. Mauá’s and Soares’ advocacy of paper currency, as already noted, was highly atypical in a world convinced of the superiority of the gold standard. In this respect, they were close to the Birmingham economists of the nineteenth century, who backed inconvertible paper money as a means to enhance employment in times of distress in their city. In a debate in the Senate, Jequitinhonha stated that Souza Franco’s opinions in favor of plurality of issue were akin to those of the Birmingham school, ‘whose theories had been proven false’. See ASI 1858, Tome I, p. 244. On the opposition of the Birmingham economists to deflation and the gold standard, Fetter (1965, pp. 177–80).

These events are examined in detail in Chaps. 2 and 3.