Appendix B: Data Sources and Methods

The investment ratios estimated in the paper were calculated by taking current price estimates of gross domestic capital formation (gross fixed investment plus changes in stocks/inventories) as a percentage of gross domestic product⁽¹⁾ at market prices. The estimates of stocks may include changes in the value of livestock.⁽²⁾ The saving ratio was estimated by adding the investment ratio to the current account to GDP or NNP ratio, measured in current prices. The saving ratio is thus defined residually, and hence incorporates all measurement error from both investment and the current account.

Australia

GDP:

GDP data from 1861-1900 are from N. G. Butlin (1962, T. 1, col. 2, p. 6), market prices, calendar years, millions of pounds. GDP data from 1900 to 1944 are from M. W. Butlin (1977, T. IV.1, col. 11, pp. 78-79), current prices, millions of dollars, converted to pounds by dividing by two. Data are for financial years (July 01 to June 30) 1900/01 to 1944/45.

Capital Formation:

Capital formation data for 1861-1900 are from N. G. Butlin (1962, T. 4, p. 16), total public and private gross domestic capital formation, market prices, calendar years, thousands of pounds. Capital formation data include changes in stocks. Changes in stocks data for 1861-1900 are from N. G. Butlin (1962, T. 7, p. 22), livestock accumulation, market prices, calendar years, thousands of pounds. Butlin (1962, p. 21) argues that "Livestock assets form a large part of the total private Australian assets, and changes in livestock assets are very large in relation to total private gross capital formation." Butlin proposed a measure of domestic capital accumulation equal to capital formation plus changes in livestock. This procedure is used for Australian data prior to 1901.

This inclusion of livestock accumulation estimates in capital formation data is often controversial, however. As Boehm (1965, p.211) argues "The alternative sequence warrants close attention in income and growth analyses because the general effects on output and expenditure of changes in livestock are not necessarily comparable with the effects of changes in fixed assets." For Australia, increases in livestock average around 1.07 percent of GDP from 1861 to 1900, and then 0.94 percent from 1900/01 to 1944/45.

From the perspective of national accounts classification, the System of National Accounts advocates the classification of livestock accumulation into estimates of changes in stocks. According to the UN Statistical Office (1952, p. 80), the value of the increase in stocks should be classified in the same way as fixed capital formation in the national accounts. The System of National Accounts was adopted by many countries in our dataset and applied retrospectively to historical data. Thus, for the purposes of this thesis, we will follow this procedure and include livestock accumulation in estimates of changes in stocks.

Capital formation data from 1900 to 1944 are from M. W. Butlin (1977, T. IV.1, pp. 78-79), private plus public fixed capital formation, current prices, millions of dollars, converted to pounds by dividing by two. Data are for financial years (July 1 to June 30) 1900/01 to 1944/45. Capital formation data include changes in stocks. Data for changes in stocks from 1900 to 1944 are from M. W. Butlin (1977, T. IV.1, pp. 78-79), current prices, millions of dollars, converted to pounds by dividing by two. Data are for financial years (July 1 to June 30) 1900/01 to 1944/45.

Current account:

Current account data for 1861-1900 are from N. G. Butlin (1962, T. 265, p. 444) for current account balance, in millions of pounds. (Number in table is defined as excess of debits, so negative of this number is used for current account balance - i.e. positive number implies surplus). Current account data for 1900-1944 are from M. W. Butlin (1977, T. IV.17, pp. 108-109) for data on exports, imports, and net property income paid overseas (cols. 1, 2, and 4 respectively). Data are in current prices, millions of dollars, converted to pounds by dividing by two. Data are for financial years (July 01 to June 30) 1900/01 to 1944/45. The value of exports less imports less net property income gives the current account balance. For the current account data prior to 1901, N. G. Butlin excluded net exports of gold and included gold production, "treating gold production as the proper current account credit of a gold producing country" Butlin (1962, p. 435). M Butlin also followed this procedure for data from 1900/01 to 1944/45. To arrive at a figure for the current account excluding all gold flows, gold production numbers are subtracted from the current account figures.

Gold:

Data on the gross value of gold production from 1861 to 1900 are from N. G. Butlin (1962, T. 55, p. 115). Data on gold production from 1900 to 1945 are from McLean (1968), pp. 83-86, row 2, gold production, millions of pounds. Data from 1901 to 1913 are for calendar years, data from 1914 are for financial years (July to June). Data on net exports of gold coin and bullion from 1861 to 1900 are from N. G. Butlin (1962), data on current account credits for gold and specie exports from T. 247, pp. 410-411, line 12, and gold and specie imports are from T. 248, pp. 413-414, line 11. Data on net exports of gold coin and bullion from 1901 to 1945 are from the Commonwealth Yearbook (various years), exports of gold bullion and specie less imports of gold bullion and specie. The Yearbook data from 1901 to 1913 are for calendar years, data from 1914 are for financial years. Data from 1901 to 1914 are converted to financial years by averaging with the data from the previous year.

The monetary gold stock includes coin and bullion held by the Treasury, mints, banks, and the public. According the the Statistical Register of New South Wales, 1876, p. 255 and 1890, p. 255, data on gold coin and bullion held by the NSW Treasury and mint for 1867-1876 and 1881-1890 suggest that mint holdings of coin are negligible (maximum value of £ 804). Bullion held at the mint reaches a maximum of £ 136, 904 in 1874. Treasury holdings of bullion and coin are listed as being negligible. Data on bullion in the hands of the public or banks are not available. The bullion component of the monetary gold stock held by the public and official agencies is therefore ignored in these calculations.

For data on the monetary gold stock of Australia, the flows of gold reported by the three mints (Sydney, Melbourne, and Perth) can be added to provide an estimate of changes in the total coin stock of the country. For New South Wales, data on gold coin in banks of issue and private hands from 1855 to 1890 are taken from the Colony of New South Wales, Statistical Register (1890, p. 234). Data for 1891 to 1899 are taken from United Kingdom, Annual Report of the Deputy Master of the Mint, various issues. Data for 1900 and 1901 are estimated by adding gold coin issued by the mint less gold coin withdrawn, plus coin imported less coin exported by the colony of New South Wales, using data from the Annual Report of the Deputy Master of the Mint (1901, p. 138).

For Victoria, there is little data available on the stock of gold coin in banks and private hands. However, the Annual Report of the Deputy Master of the Mint (1893, p. 123; 1899, p. 123; 1901, p. 138) gives data on gold coin flows for the colony of Victoria from the opening of the Melbourne mint in 1872. The sum of the data on gold coin issued less coin withdrawn, plus gold coin imported less coin exported from the colony of Victoria gives the change in gold coin in the colony.

For Western Australia, the Annual Report of the Deputy Master of the Mint (1901, p. 138) gives data on gold coin flows for the colony of Western Australia from the opening of the Perth mint in 1899. The sum of the data on gold coin issued less coin withdrawn, plus gold coin imported less coin exported from the colony gives the change in gold coin in Western Australia.

Data on the stock of gold bullion held by the banks is taken from S. J. Butlin et al. (1971), Australian Banking and Monetary Statistics 1817-1945, T. 1, p. 113, average of weekly figures for December quarter until 1900, then average of weekly figures for June quarter. The data are for total bullion, but according to the notes on p. 77, bank bullion holdings are mainly gold bullion.

Changes in the monetary gold stock for Australia are then estimated as the sum of changes in the stock of gold coin in banks of issue and private hands in New South Wales, plus the sum of changes in the gold coin stock in the colonies of Victoria and Western Australia, plus the change in bullion held by Australian Trading Banks. Data from 1861 to 1900 are for calendar years. Data on the monetary gold stock for 1901 to 1944 are taken from Australian Banking and Monetary Statistics 1817-1945, T. 42, pp. 453-457, total gold coin held by banks and the public, and bullion held by Australian Trading banks, T. 1, pp.115, end of financial year (June quarter averages). An estimate of the gold stock for the financial year 1899 was calculated as the average of the gold stock in 1899 and 1900. The estimated gold stock at the end of 1899 and 1900 was calculated by taking the cumulative sum of net additions to coin in the colonies of Victoria and Western Australia from 1872 to 1899 and 1900 (ie the sum of gold coin issued less coin withdrawn, plus gold coin imported less coin exported), plus the stock of gold coin in banks of issue and private hands in New South Wales in 1899 and 1900, plus the gold bullion holdings of Australian Trading Banks. Data on changes in the monetary gold stock from 1900/1901 onwards are calculated as the first difference of the sum of coin stock held by banks and the public plus bullion held by trading banks, listed in Australian Banking and Monetary Statistics 1817-1945.

The main drawback with estimating the monetary gold stock using this method is that it tends to overstate the actual gold stock. As S. J. Butlin et al. (1971, p.92) caution: "Because of inadequacies in the gold production figures and the freedom until 1914 with which coins in circulation in Australia were imported and exported and the probable statistical significance thereafter of unrecorded imports and exports of coins, it is not possible to produce satisfactory estimates of coinage in use in the nineteenth century, or to be more exact, it would not be possible to do so without conducting much more detailed research." Another relevant comment on the accuracy of the statistics is made in the Annual Report of the Deputy Master of the Mint, 1902, p. 138: The above return shows that only 13.73 percent of the gold coined at the Melbourne, Sydney, and Perth Mints during the last 29 years has been retained in the States coining it. The amount actually retained is probably much less than this, for considerable quantities are taken away by passengers for Europe which do not appear in the Customs House Returns, and which probably are not counterbalanced by sums brought in by incoming passengers... Thus in Victoria, the amount of all coined metals held by Banks on the 31st December 1901 was only £3 928 107 more than they held on the 31st December 1872, or only 26.84 percent of the amount apparently retained in the State since that date, and in a State where the practice of keeping Bank accounts is almost universal, and the circulation of L1 Bank Notes very large ... it is not probable that the amount of gold coin in private hands is very large." On a compounded basis, this is equivalent to overstating outflows by 4.43% per year i.e. (10.0443)²⁹ = 0.2684. Some of the gold may have also made its way into banks in other states instead of being unrecorded flows. This is equivalent to overstating the change in the monetary gold stock by less than 5% per year. Since reliable data are not available on funds taken overseas and inbound by passengers, there is little that can be done to the estimated figures for the monetary gold stock to adjust for this source of error.

Canada

GDP:

GDP data from 1870 to 1926 are calculated by subtracting net interest and dividend payments from abroad from the figures for GNP. Data for GNP are from from Urquhart (1986, T. 2.1, pp. 11-15, row 25), market prices, thousands of dollars. GNP data from 1927 to 1944 are from Urquhart and Buckley (1965, series E 27, p. 131), gross national expenditure at market prices, millions of dollars. Data for net interest and dividend payments from 1870 to 1926 are from Urquhart (1986, T. 2.1, pp. 11-15, row 8 less row 21), interest and dividends credits less interest and dividends debits, market prices, thousands of dollars. Data from 1927 to 1944 are from Urquhart and Buckley (1965, series F 60 less F66, p. 160), current receipts of interest and dividends less current payments of interest and dividends, millions of dollars.

Capital Formation:

Capital formation data for 1870 to 1926 are from Urquhart (1986, T. 2.2, pp. 16-17, col. 8), grand total, gross fixed capital formation, current dollars, millions. Data from 1927 to 1944 are from Urquhart and Buckley (1965, series E 17, p. 131), total business gross fixed capital formation, millions of dollars. Capital formation data include changes in stocks. Changes in stocks data available only from 1927 to 1944, from Urquhart and Buckley (1965, series E 21, p. 131), total value of the physical change in inventories, millions of dollars.

Current account:

Current account data from 1870 to 1926 are from Urquhart (1986, T. 2.4, pp. 20-25, rows 9 and 22), total current credits and total current debits, thousands of dollars. Data include net exports of gold coin and bullion. Data from 1927 to 1944 are from Urquhart and Buckley (1965, series F 71, p. 160), net balance, all countries, millions of dollars. Data include net exports of non-monetary gold. Data for the current account excluding all gold flows from 1870 to 1926 are calculated by subtracting net exports of gold coin and bullion from the current account. Data on the current account excluding gold from 1927 to 1944 are calculated by subtracting net exports of non-monetary gold from the current account.

Gold:

Data on net exports of gold coin and bullion for 1870 to 1926 are from Urquhart (1986, T. 2.4, pp. 20-25, rows 2 and 16), exports of gold coin and bullion, less imports of gold coin and bullion, thousands of dollars. Data on net exports of non-monetary gold for 1868 to 1899 are from Rich (1988, T. A-3, pp. 245-246, col. 3), millions of dollars. Data on net exports of non-monetary gold for 1900 to 1913 are calculated by taking the sum of net gold exports and the change in the monetary gold stock. Data on net gold exports are from Urquhart (1986) as listed above. Data on the change in the monetary gold stock are from Rich (1988) as listed below. Data on net exports of non-monetary gold from 1927 to 1944 are from Urquhart and Buckley (1965, series F 58, p. 160), current receipts, net exports of non-monetary gold, millions of dollars, all countries. Data on net exports of monetary gold from 1927 to 1937 are from Urquhart and Buckley (1965, series F99, p. 164), monetary gold movement (net). Data for 1938-1939 are from calculated by using the change in the gold holdings of the Bank of Canada, from Canada Yearbook (1939, p. 934) and (1941, p. 805). Data for 1940-1945 are from Dominion Bureau of Statistics (1949), Statement 8 - Canada's holdings of gold and US dollars, p. 57, converted to Canadian dollars at the official exchange rate of US$ 0.909090 = C$ 1. Monetary gold movements are taken as the negative of the change in the Canadian gold stock. Data on net exports of gold from 1927 to 1944 are calculated as the sum of net exports of non-monetary gold minus the change in the monetary gold stock.

Data on the monetary gold stock from 1869 to 1871 are calculated as the sum of bank gold and subsidiary coin plus the stock of gold held against Dominion notes. Data on bank gold and subsidiary coin are from Curtis (1931, p.36) current gold and subsidiary coin, thousands of dollars, December figures. Data on gold held against Dominion notes are from Curtis (1931, p. 92), thousands of dollars, December figures. Data on the monetary gold stock from 1872 to 1913 from Rich (1988, T. A1, pp. 239-242), Dominion government gold holdings plus the gold holdings of chartered banks, end of year. Data on the monetary gold stock from 1914 to 1926 are calculated as the sum of official Canadian gold reserves, plus bank gold and subsidiary coin holdings, plus bank gold held in the central reserves of the banking system (a private gold reserve managed by trustees from the Ministry of Finance and the Canadian Bankers Association). Data on the offical gold reserves are from the Canada Yearbook 1927-28, p. 857, T. 3, total Canadian gold reserves, end of year figure. Data on bank gold and subsidiary coin are from Curtis (1931, p.36) current gold and subsidiary coin, thousands of dollars, December figures. Data on bank gold held in the central gold reserves are from Curtis (1931, p. 35), gold coin, thousands of dollars, December figures.

Denmark

GDP:

GDP data from 1850 to 1944 are from Hansen (1977, T. 5, pp. 261-263), gross factor incomes (bruttofaktorindkomst) plus net import of goods and services (netto import af varer og tjenester) plus indirect taxes less subsidies (indirekte afgifter - pristilskud), equalling disposal of merchandise and services (varer og tjenester til radighed), current market prices (lobende markedspriser), millions of kroner.

Capital Formation:

Capital formation data from 1850 to 1944 are from Hansen (1977, T. 6, pp. 264-266, col. 2), gross investment (brutto-investering), current prices, millions of kronor. Data do not include stocks.

Current Account:

Current account data from 1874 to 1944 are from Bjerke and Ussing (1958, T. VI, pp. 152-153, col. 4), net imports of goods and services, millions of kroner.

Gold:

Data on the monetary gold stock from 1874 to 1906 are from the U.S. Annual Report of the Director of the Mint (various years). Data on the gold stock for 1875 and 1894 are calculated by adding net exports data for the following year to gold stock figures for the following year. Data on the gold stock for 1877, 1886-1888 are calculated by subtracting net exports data for the year from the gold stock figures for the previous year. Data on the gold stock for 1874 and 1878 are calculated as the average of the preceding and following years. US dollar values are converted to crowns at the exchange rate of 0.268 US dollars per crown. Data on the gold holdings of the central bank (Danmarks Nationalbank) from 1907 to 1944 are from Statistisk Arbog Danmark (Statistical Yearbook) (various years). The change in the monetary gold stock is assumed to equal the change in the gold holdings of the central bank.

Data on gold exports and imports for 1874 to 1906 are from the U.S. Annual Report of the Director of the Mint (various years). Data on gold imports and exports from 1910 to 1930 are from the League of Nations Memorandum on the Balance of Payments (1924, 1927, 1931, 1932), bullion and specie import, bullion and specie export, then gold import and export. The figures for bullion and specie import include some non-gold coin. However, there is a close correspondence between gold bullion and specie trade and total bullion and specie trade from the League of Nations figures for the years with overlapping data (1922-1930), suggesting that the non-gold element of bullion and specie trade was small. On this basis, the figures for total bullion and specie trade were used from 1910 to 1921, then gold bullion and gold specie trade figures were used from 1922 to 1930. Data on the net export of gold from 1907 to 1909 and from 1931 to 1934 and are assumed to be equal to the negative of the change in the monetary gold stock (ie an increase in the monetary gold stock implies and equal amount of gold imports). According to the League of Nations Memorandum on Balance of Payments and Foreign Trade Balances 1911-1925, vol. II, p. 85, Danish trade statistics exclude gold bullion and all specie. Silver bullion is included in merchandise trade. Thus the current account data should not include gold trade. Data on monetary gold exports for 1935 to 1944 are from Statistisk Arbog Danmark (Statistical Yearbook) (various years). Data on net exports of gold for 1935 to 1944 are assumed to equal the monetary gold exports.

Finland

Finland became an autonomous Grand Duchy connected with Russia in 1809, and declared independence from Russia on December 6, 1917. Finland had its own monetary system from 1860, and was on the gold standard from 1877 to 1914. The unit of currency was the Finnmark (Finnish markka), equivalent to one French gold franc. Finland returned to the gold standard on January 1, 1926, and remained on the gold standard until October 1931.

GDP:

GDP data from 1860 to 1944 from from Hjerppe (1989, T. 3A1, col. 1, pp. 201-203 ), market prices, thousands of FIM.

Capital Formation:

Capital formation data from 1860 to 1938 from from Hjerppe (1989, T. 3A1, col. 5, pp. 201-203), gross fixed capital expenditure, market prices, thousands of FIM. Changes in stocks data from 1860 to 1938 from from Hjerppe (1989, T. 3A1, col. 7, pp. 201-203), increase in stocks + statistical discrepancy, market prices, thousands of FIM.

Current Account:

Current account data from 1860 to 1938 from from Hjerppe (1989, T. 3A1, cols. 2 and 6, pp. 201-203), imports of goods and exports of goods, market prices, thousands of FIM. The statistical discrepancy includes data on net exports of services.

Gold:

Net Exports of gold from 1860 to 1909 from Pihkala (1970, T. 2, pp. 80-91), value of exports, 43: Raha (coinage), 1000 mk, less Pihkala (1970, T. 7, pp. 112-123), value of imports, 1252: Hopea, kulta, platina (silver, gold, platinum), 1000 mk. This data includes some silver and platinum imports and exports. Net exports of gold from 1910 to 1930 are from the League of Nations Memorandum on the Balance of Payments (1924, 1927, 1931, 1932), gold bullion and specie import, less gold bullion and specie export. According to the League of Nations (1927, p. 104), Finnish merchandise trade statistics include gold specie and bullion movements in the data for manufactured gold and unwrought gold, respectively. Net exports of gold from 1931 to 1944 are from Ulkomaankauppa Vuosijulkaisu (various years), item numbers 951-953, precious metal imports: Kultaa: valmistamatonta ja jatteita, lankaa ja levya, teoksia muunlaisia, rahaa, muita (gold, manufactured and unmanufactured, coin), then from 1939 items 61-005, 61-102 to 61-014, and 62-001. Exports data are given in series 61-002 Kultaa: valmistamaton seka jatteet ja romu and 62-001 Kutaraha (gold coin).

Data on the monetary gold stock from 1871 to 1944 are taken from Suomen tilastollinen vuosikirja (Statistical yearbook of Finland), various years, gold assets of the Bank of Finland. The change in the monetary gold stock is calculated as the first difference of the gold assets of the Bank of Finland. Data on the current account excluding gold are calculated by subtracting net gold exports (which include some silver and platinum) from the current account data.

France

GDP:

GDP data from 1850 to 1900 are from Levy-Leboyer and Bourguignon (1985, T. A-III, series 1, pp. 329-332), produit interieur brut, millions de francs courants. Data from 1901 to 1944 are from Villa (1983, p. 459, series PIBQ), Production Interieure Brute en valuer - en gros franc courants.

Capital Formation:

Capital formation data from 1850 to 1900 are equal to gross fixed capital formation plus changes in stocks. Capital formation data are from Levy-Leboyer and Bourguignon (1985, T. A-III, series 4, pp. 329-332), investissements, bruts, (gross fixed capital formation), millions de francs courants. Changes in stocks data from 1850 to 1900 are from Levy-Leboyer and Bourguignon (1985, T. A-III, series 8, pp. 329-332), variations des stocks (change in stocks), millions de francs courants. Capital formation data from 1901 to 1944 are are equal to gross fixed capital formation plus changes in stocks. Capital formation data are from Villa (1983, p. 439, series IE, IG, IM), Investissement des Entreprises, Investissement des Administrations, Investissement des Menages, en valeur, en Gros franc courants, (sum of investment by businesses, government, households). Changes in stocks data are from Villa (1983, p. 457, series DS), variations de stocks en valeur - en Gros franc courants (changes in stocks).

Current Account:

Current account data from 1850 to 1900 are from Levy-Leboyer and Bourguignon (1985, T. A-III, series 5, 6, and 7, pp. 329-332), exportations (exports) plus gains invisibles (invisible earnings) minus importations (imports), millions de francs courants. Current account data from 1901 to 1923 are from Villa (1983, p. 437: series EXPORT, plus p. 448: series SUS, plus p. 438: series IDVX, plus p. 435: series DREX, plus p. 435: series DOMX, minus p. 439: series IMPORT minus p. 444: series ODRX), exportations en valeur (exports), plus solde des utilisations de services (balance of services), plus interets et dividendes verses par l'exterieur (interest and dividends), plus depenses et recettes exterieurs (external revenue and expenditure), plus dommages de guerre verses par l'exterieur (war reparations), minus importations en valeur (imports), minus operations diverses de rapartitions exterieur (sundry external transactions), all en Gros franc courants. Current account data from 1924 to 1944 are from Villa (1983, p. 436, series EBX), epargne brute de l'exterieur - en gros franc courants.

Gold:

Data on net exports of gold and silver for 1850 to 1875 are from the US Annual Report of the Mint (1900, pp. 424-425). US dollar values are converted to French Francs at the exchange rate of 0.193 US dollars per Franc.

Data on net exports of gold from 1876 to 1914 from NBER Macrohistory database (series 14114), excess of gold exports over imports, sum of 12 months data. Data for 1915 to 1930 are from the League of Nations (1924, 1927, 1931, 1932), gold bullion and specie exports less gold bullion and specie imports. Data from 1931 to 1944 are from Annuaire Statistique (1966, Ch. 40, T. I, p. 365), mouvements d'or, million de francs 1928. The figures for net exports of gold include gold and silver from 1850 to 1870.

Data on changes in the monetary gold stock for 1851 to 1865 are calculated from Annuaire Statistique (1966, T. III, p. 516), Banque de France encaisse d'or et d'argent (Bank of France holdings of gold and silver, average for year). Changes in the monetary gold stock are set equal to the change in gold and silver holdings of the Bank of France. Data from 1866 to 1870 are from NBER Macrohistory database (series 14100), Bank of France metallic reserve (gold and silver), December figures. Data from 1871 to 1913 are from Saint Marc (1983, pp. 24, 25, col. E), Monnaies d'or, total existant evalue (total gold existing in country). Data from 1913 to 1927 are from Annuaire Statistique (1966, T. III, p. 517), Banque de France encaisse d'or (Bank of France gold holdings, annual average). Data from 1928 to 1945 are from Annuaire Statistique (1966, T. II, p. 562), Banque de France encaisse d'or (Bank of France gold holdings, year end), multiplied by the price of gold from Annuaire Statistique (1966, T. II, p. 562), cours de l'or a Paris, cours d'achat par la Banque de France. Price data from 1938 to 1945 are from Annuaire Statistique (1952, p. 503), cours de l'or a Paris et a Londres depuis 1938, end of December figures.

According to Annuaire Statistique (1952, p. 195), external trade data excludes gold, silver, and copper coin, raw gold and silver, gold and silver in bars, ingots, and powder. On this basis, the current account data excluding gold were assumed to equal the original current account data.

Germany

GDP:

GDP data from 1850 to 1939 are from Hoffman (1965, T. 248, col. 5, pp. 825-826), Nettosozialprodukt zu Markt-preisen, (NNP at market prices), mill mark, raised 8.4% to approximate GDP following procedure in Maddison (1991).

Capital Formation:

Capital formation data from 1850 to 1939 from Hoffman (1965, T. 248, col. 2, pp. 825-826), Nettoinvestitionen (net investment), mill mark. Data includes stocks. Stocks data from 1851 to 1939 are from Hoffman (1965, T. 32, col. 4, p. 237), Investitionen der Landwirtschaft, Vorrate (investment in agriculture, stocks), plus T. 36, col. 3, p. 247, Investitionen im Gewerbe, Vorrate, (investment in trade/industry, stocks). Data from 1851 to 1913 are for the sum of Anlagen and Vorrate (fixed assets and stocks). Data for 1851 to 1913 estimated by using the average of Vorrate (stocks) to the sum of Anlagen and Vorrate for the years 1924-1959, and multiplying by the annual sum data.

Current Account:

Current account data from 1860 to 1938 are from Hoffman (1965, pp. 825-826, col. 4), Saldo der Leistungsbilanz. The current account data exclude net exports of precious metals. The current account excluding gold is calculated by adding net exports of precious metal to the current account and subtracting net exports of gold.

Gold:

Exports data from 1872 to 1942 from Germany, Deutsche Bundesbank (1976), T. J 1.03, p. 324, col. 11, Ausfuhr, darunter gold (exports of gold) and col. 13, Einfuhr, darunter gold (imports of gold). Exports data for 1872 to 1879 are estimated by taking the average of gold exports to total exports of precious metal for the period 1880 to 1913 and multiplying by the figure for corresponding figure for total precious metal exports. Imports data for 1872 to 1875 are estimated by taking the average of gold imports to total imports of precious metal for the period 1876 to 1913 and multiplying by the figure for corresponding figure for total precious metal imports.

Gold holdings of the Reichsbank for 1876-1945 from Germany, Deutsche Bundesbank (1976), T. C 1.01, p. 36, col. 2, Aktiva: Gold in Barren and Munzen (gold in bars and coin). Data for 1878, 1879, 1883, 1887, 1888, 1890 are not shown separately, but data are given for sum of Gold in Barren und Munzen and Deutsche Scheidemunzen (total gold in bars and coin and German subsidiary coin), Data for 1878, 1879, 1883, 1887, 1888, 1890 are estimated by taking the ratio for the previous year of gold in bars and coin to the sum of gold in bars and coin plus subsidiary coin, and multiplying by the corresponding sum for the missing year. Gold coin in circulation from 1876 to 1913 from Deutsche Bundesbank (1976), T. B 1.01, p. 14, Munzen, Goldmunzen (coin, gold coin), col. 7. The monetary gold stock is calculated as the sum of gold coin in circulation plust the gold holdings of the Reichsbank.

Italy

GDP:

GDP data are calculated by taking the figures for GNP and subtracting net factor incomes and current transfers received from the rest of the world. GNP data from 1861 to 1925 are from Italy, ISTAT (1957, T. 36, col. 6, p. 249-250), reddito nazionale lordo ai prezzi di mercato (gross national product at market prices), milioni di lire. Net factor incomes from abroad are from ISTAT (1957, T. 36, col. 3, p. 249-250), Esterno redditi netti dall'estero, milioni di lire. Current transfers are from ISTAT (1957, T. 39, col. 4, p. 255), trasferimenti correnti, saldo, milioni di lire. GNP data for 1926 to 1945 are from ISTAT (1986), T. 8.1, p. 143, col. 1, reddito nazionale netto (net national income), plus col. 2, ammortamenti (depreciation). Net factor incomes from abroad are from ISTAT (1986, T. 8.11, p. 151, cols. 2, 7), exportazione redditi dei fattori less importazione redditi dei fattori (exports of factor income less imports of factor income).

Capital Formation:

Data from 1861 to 1925 are from ISTAT (1957, T. 44, col. 6, pp. 264-265), investimenti lordi (gross investment), current prices. Data includes stocks. Data for 1926 to 1945 are from ISTAT (1986), T. 8.1, p. 143, col. 6, investimenti lordi (gross investment). Data on changes in stocks from 1861 to 1925 are from ISTAT (1957, T. 44, pp. 264-265, col. 5), Variazioni scorte (change in stocks). Data on changes in stocks for 1926 to 1945 are from ISTAT (1986, T. 8.28, p. 166, col. 5), variazione delle scorte (change in stocks).

Current Account:

Data from 1861 to 1925 are from ISTAT (1957, T. 39, col. 5, p. 255), Bilancia dei pagamenti correnti, conto transazioni e trasferimenti correnti (balance of current payments, including current transfers) Data for 1926 to 1945 are from ISTAT (1986, T. 8.11, p. 151, col. 11), Saldi, transazioni correnti (settlement of current transactions), equal to the sum of net exports of goods, services, factor income, and transfers. Merchandise trade figures include silver bullion, but exclude gold bullion and silver coin of the Latin Union. The current account excluding gold is therefore the same as the original current account figure.

Gold:

Data for 1861 to 1936 are from Di Mattia (1967). Data for gold and silver coin withdrawn from T. 8, pp. 473-474, cols. 1 and 2, Retiri di monete, oro and argento ed eroso misto (withdrawn coins, gold and silver). Data for gold and silver coined and recoined from T. 9, pp. 477-478, col. 1, Coniazioni e Reconiazioni di Monete, Oro and Argento (coinage and recoinage, gold and silver). Imports and exports data from T. 10, pp. 481-482, cols. 1, 2, 5, 6, Esportazioni oro grezzo (exports unrefined gold), Esportazioni oro monetato (exports gold coin), Importazioni oro grezzo (imports unrefined gold), Importazioni oro monetato (imports gold coin). Data for 1861 to 1877 for gold imports and exports include silver coin & unrefined silver.

Data for 1861 to 1877 for gold specie exports are calculated by taking the average of gold specie to total specie exports for the years 1878-1936, and multiplying by the annual figure for total specie exports. Data for unrefined gold exports are calculated by taking the average of gold specie to total specie exports for the years 1878-1936, and multiplying by the annual figure for total unrefined gold and silver exports. The data for gold imports are calculated similarly.

The change in the monetary gold stock is calculated as the sum of specie imports less specie exports plus coinage and recoinage less withdrawn coin. Data on official reserves of gold from ISTAT (1968, T. 83, p. 105, col. 11), Riserve Ufficiali, Di Cui Oro, lire milioni (official reserves of gold coin, millions lire).

For the period from 1862 to 1866 when Italy was on a bimetallic standard, the change in the monetary stock includes coinage and recoinage and net specie imports of silver as well as gold. The data on net exports also include net exports of silver bullion and specie for this period.

Japan

The New Coinage Act of 1871 declared the gold yen as the standard unit of value and legal tender for transactions of any value. Silver coins were relegated to subsidiary money, legal tender up to 10 yen. However, the Act also declared the silver Yen Trade Dollar as legal tender within the confines of treaty ports. An amendment in May 1878 made the silver Trade Dollar legal tender throughout the Empire of Japan. Thus both gold and silver were legal tender within Japan and for all foreign transactions from 1878 to 1897. It wasn't until the Coinage Act of 1897 declared the gold yen as the standard unit of value and legal tender that Japan officially adopted the gold standard. The coinage of the Yen Trade Dollar ceased, and they were gradually withdrawn from circulation.

GDP:

GDP data are calculated by taking the figures for GNP and subtracting net factor incomes and transfers received from the rest of the world. GNP data from 1885 to 1929 are from Ohkawa et al (1979, T. A1, pp. 251-253, col. 7), gross national expenditure at market prices, millions of yen, current prices. GNP data from 1930 to 1944 are from Ohkawa et al (1979, T. A2, p. 254, col. 7), gross national expenditure at market prices, millions of yen, current prices. Data on net factor incomes received from the rest of the world are from Ohkawa et al (1979, T. A31, pp. 332-335, cols. 3, 6), exports: income from abroad less imports: income from abroad. Data on net transfers from abroad are from Ohkawa et al (1979, T. A31, pp. 332-335, col. 8), net transfers from abroad.

Capital Formation:

Capital formation data equal gross domestic fixed capital formation plus changes in stocks. Data from 1885 to 1929 from Ohkawa et al (1979, T. A1, pp. 251-253, col. 3), gross domestic fixed capital formation. Data from 1930 to 1944 from Ohkawa et al (1979), T. A2, p. 254, col. 3, gross domestic fixed capital formation. Change in stocks data from 1885 to 1929 are calculated by taking Inventory Data as % GNE taken from p. 63 of Ohkawa et al (1979) quoting from Fujino, Shozaburo and Akiyama, Ryoko (1973), Zaiko to Zaiko Toshi, 1880-1940, Hitotsubashi Daigaku, Keizai Kenkyujo, and multiplying by the figure for GNE. Inventories data from 1930 to 1944 are from Ohkawa et al (1979, T. A2, p. 254, col. 4), increase in stocks.

Current Account:

Data for 1868 to 1940 from Ohkawa et al (1979, T. A31, pp. 332-335, col. 9), surplus on current account (excluding reparations). Data does not include nonmonetary gold or monetary gold shipments. Data from 1940 to 1944 from Ohkawa et al (1979, T. A32, p. 336, col. 9) surplus on current account (excluding reparations).

Gold:

Gold coinage data for 1871 to 1897 are from Matsukata (1899), T. II, p. 13, Amount of gold coins issued, and T. III, p. 14, Amount of silver coins issued. Data for 1900 to 1939 are from Japan, Ministry of Finance, Financial Annual of Japan, nos. 1, 10,16, 26, 36, 40, coins turned out by the mint. Data for 1913 to 1936 on coinage withdrawn are from Japan, Bank of Japan (1932, 1937), Economic Statistics of Japan, p. 2, amount of coin melted by the mint.

Data on gold exports and imports for 1872-1933 on exports and imports of gold from Ishibashi (1935), details of coin and bullion exported, pp. 431-433, details of coin and bullion imported, pp. 436-437, gold bullion and total of gold coin and bullion. Data are for Japan Proper (ie excludes Korea and Taiwan after annexation). Data for 1934-1936 on net exports of gold coin and bullion from Japan, Japan Statistical Yearbook (1949, T. 280, pp. 520-521, col. 7-8), domestic (Japan Proper) exports and imports of gold coin and bullion. Silver exports and imports data for 1872-1933 are from Ishibashi (1935), details of coin and bullion exported, pp. 433-435, details of coin and bullion imported, pp. 437-439, silver bullion and total of silver coin and bullion. Data are for Japan proper (ie excludes Korea and Taiwan). Data for 1934-1936 on net exports of silver coin and bullion from Japan Statistical Yearbook (1949), T. 280, pp. 520-521, col. 7-8, domestic (Japan Proper) exports and imports of silver coin and bullion. Data on gold coin and bullion imports and exports and silver coin and bullion imports and exports for 1937 to 1945 are from Japan, Hundred Year Statistics of the Japanese Economy (p. 561, and Supplement, p. 157)

Data on coins existing in the country from 1868 to 1900 from Japan, Financial Annual of Japan (1901). Data on coins existing in the country from 1901 to 1914 from Shinjo (1962), T. XXB, p. 101. Data on the estimated stock of specie in the country from 1872 to 1914 is derived from the data on coins existing in the country. Data after 1914 take the previous years estimate of specie existing in the country, plus coins turned out by the mint, less coin melted by the mint, less net exports of specie. Changes in the monetary stock take the change in estimated stock of specie, less net exports of bullion. Data on the monetary stock for 1878 to 1896 include silver and gold. Data after 1897 include gold only. Data after 1934 are based on coinage less recoinage less net exports of gold coin and bullion.

Norway

GDP:

GDP data from 1865 to 1938 are from Norway, Statistisk Sentralbyra (1965, T. 49, r. 11, pp. 340-343), current prices, millions of kroner.

Capital formation:

Capital formation data from 1865 to 1938 are from Norway, Statistisk Sentralbyra (1965, T. 49, r. 5, pp. 340-343), current prices, millions of kroner. Data include increase in stocks from 1909. Data on increase in stocks for 1900-1908 are from Statistisk Sentralbyra (1953, T. 2, p. 107, col. 9), Lagerendring (change in stocks). Data for 1909-1939 are from Statistisk Sentralbyra (1965, T. 49, pp. 340-343, r. 7). Increase in stocks not estimated for the years 1865-1899. For 1900-1913 and 1921-1929 only net increase in standing forests and in livestock are included. From 1930 increase in standing forestsis regarded as fixed capital investment.

Current Account:

Current account data from 1865 to 1899 are from Mitchell (1993, T. J3, pp. 922, 927), current prices, millions of kroner. Current account data for 1900-1929 are from Statistisk Sentralbyra (1953, T. 12, p. 126, col. 5), Netto oking eller nedgang i Norges netto fordringer pa utlandet (Net increase or decrease in Norway's foreign assets). Data for 1930-1939 from Statistisk Sentralbyra (1965, T. 25, pp. 184-185, r. 16). The current account data exclude crude gold and silver and coins.

Gold:

Data on gold and silver holdings from 1865-1913 are from Norway, Statistisk Sentralbyra (1978, T. 257, col. 1&6, p. 484), Metallfondene (gold and silver) at bank of Norway. Data from 1914-1939 from Statistiske Oversikter (1948, T. 159, rows. 1 & 4, pp. 300-303), Gullbeholdning (gold stock), and Midlertidig anbrakt i gull (temporarily invested in gold). The change in the monetary gold stock is calculated as the first difference of the gold stock at the Bank of Norway (gold and silver stock until 1914). According to United States, Report of the Director of the Mint (1886, p. 222), "the amount of gold in banks, other than the Bank of Norway, or in circulation, has probably not been considerable."

Data on gold exports and imports for 1865 to 1896 are calculated as the first difference of the gold and silver holdings of the Bank of Norway (the negative of the change in the monetary stock equals the net exports - i.e. an increase in the monetary stock corresponds to an import). Data on gold exports and imports from 1895 to 1909 and for 1931 to 1944 are from Norway, Norges Handel (annual issues), unwrought platinum, gold and silver and coins and medals. Missing trade returns data for 1897 and 1899 are interpolated linearly from the previous and subsequent years. Data on net exports of gold from 1910 to 1930 are from the League of Nations (1927, 1931, 1932), exports of bullion and specie less imports of bullion and specie (including silver and platinum, as per the trade returns).

Russia

GDP:

Data on net national product are from Gregory (1982, T. 3.2), net national product, Russian Empire, millions of credit roubles.

Capital Formation:

Data on capital formation are from Gregory (1982, T. 3.2), net investment, Russian Empire, millions of credit roubles. Data include inventories. Data on inventories are from Gregory (1982, T. 3.2), inventories, total, Russian Empire, millions of credit roubles.

Current Account:

Data on the current account are from Gregory (1982, T. 3.2), net foreign investment, Russian Empire, millions of credit roubles. The current account data include net exports of silver. Data on the current account excluding gold and silver are calculated by subtracting the figures for net silver exports from the figures for net foreign investment for the years 1886 to 1896. Net silver exports data are from Gregory (1982, T. M1, p. 314, col. 2 less col. 5), silver exports less silver imports, millions of credit roubles.

Data on the monetary gold stock are from United States, Annual Report of the Director of the Mint (various years). Data for 1880-1891 are calculated as the sum of bullion and gold coin of the Treasury at the Bank of Russia plus gold specie on hand at the Bank of Russia (belonging especially to the Bank of Russia). Data on the monetary gold stock from 1891 to 1914 are the sum of gold coin in circulation plus gold coin and bullion in the treasuries and the State Bank. Data on the silver stock for 1885 to 1897 are from United States, Annual Report of the Director of the Mint (various years). Data on the silver stock for 1886-1891 are calculated by adding coingage less recoinage less net exports of silver to the stock of silver from the previous year. Data on the change in the monetary gold stock include changes in the silver stock for the years 1886 to 1896.

Data on net exports of gold from 1885 to 1913 are from United States, Annual Report of the Director of the Mint (various years). Data for 1900, 1901, 1912 are calculated as the negative of the change in the monetary gold stock. Data on net exports of silver from 1885 to 1899 are from United States, Annual Report of the Director of the Mint (various years). Data on net exports of gold include net exports of silver from 1886 to 1896.

US dollar values are converted to roubles at the exchange rate of 0.7718 roubles per US dollar until the end of 1897, then 0.514556. Data on monetary gold and silver, net exports of gold and silver are converted from roubles to credit roubles by multiplying by a factor of 1.5.

Sweden

GDP:

Data on gross domestic product at factor cost data for 1861 to 1945 are from Krantz and Nilsson (1975, T. 1:2, pp. 154-155, col. 4). Indirect taxes and customs duties from Krantz and Nilsson (1975, T. 1:2, pp. 154-155, col. 3). GDP at market prices calculated as sum of GDP at factor cost plus indirect taxes and customs duties.

Capital Formation:

Data on capital formation for 1861 to 1945 are from Krantz and Nilsson (1975, T. 1.1, col. 6, pp. 150-152), domestic investment. Data does not include stocks. Data on stocks for 1861 to 1945 are from Johansson (1967, T. 1, col. 3, pp. 38-39), changes in livestock. Data on capital formation include domestic investment plus changes in livestock.

Current Account:

Data on the current account from 1861 to 1930 from Lindahl et al (1937), T. 174, col. 8, pp. 598-599. Net balance on goods and services. Data do not include shipments of gold and silver. Data for 1931 to 1935 from Ohlsson (1969), T. B:1., p. 123, col. 6, bytesbalansens saldo. Data does not include net exports of gold and silver. Data for 1936 to 1945 from Sweden, Historisk Statistik for Sverige (1960), T. 33, p. 64, r. 6. Data does not include net exports of gold and silver. Net exports of silver are added to the current account balance.

Gold:

Data on net exports of gold from 1861 to 1874 from Lindahl et al (1937, T. 175, pp. 604-605, col. 1), net imports of gold according to trade statistics. Data on gold and specie flows from 1861 to 1871 included both gold and silver. According to Lindahl et al (1937, p. 610), the figures included only insignificant quantities of gold, so the net export of gold is assumed to be zero from 1861-1871. Data from 1875 to 1913 from Lindahl et al (1937, T. 175, pp. 604-605, col. 4), change in bank holdings plus gold absorbed by industry. Lindahl et al argue that the trade statistics data suffer from several shortcomings with regards to the net exports of gold, and use the estimated change in the stock of gold held by banks and the value of gold absorbed by industry as the measure of net gold exports. Data from 1914 to 1945 are from Sweden, Historisk Statistisk for Sverige (1960, T. 30, p. 62, cols. 2, 5, 9, 12), imports of unmanufactured gold and gold coins and exports of unmanufactured gold and gold coins. Data on changes in the monetary gold stock for 1875 from Lindahl et al (1937, T. 175, col. 2, p. 604), increase in bank holdings of gold. Data on the monetary gold stock from 1876 to 1945 from Historisk Statistik for Sverige (1960, T. 76, pp. 97 and T. 77, p. 98), gold holdings of banks (including Riksbank). Change in the monetary gold stock is calcuated as the change in the gold holdings of banks.

United Kingdom

GDP:

From 1850 to 1969 from Mitchell (1988), Ch. 16, T. 5, pp. 831-832, Gross domestic product at market prices. Data from 1870 to 1944 from Feinstein (1972), T. 3, pp. T10-T11, gross domestic product at market prices.

Capital Formation:

Data from 1850 to 1869 from Mitchell (1988), ch 16, T. 5, pp. 831-832, gross domestic fixed capital formation plus value of physical increase in stocks. Data from 1870 to 1920 from Feinstein (1988), T. 17, pp. 462-463, gross domestic fixed capital formation plus value of physical increase in stocks and works in progress. Data from 1921 to 1944 from Feinstein (1972), t. 2, pp. T8-T9, gross domestic fixed capital formation plus value of physical increase in stocks and works in progress.

Current Account:

Data from 1850 to 1869 from Imlah (1958), T. 4, pp. 70-72, balance on current account. Data includes net export of gold and silver bullion and specie. Data from 1870 to 1920 from Feinstein (1988), T. 17, pp. 462-463, net investment abroad. Data includes balance of payments surplus on current account, less net imports of all gold and silver, plus net increase in holdings of monetary gold and silver. Data from 1921 to 1944 are from Feinstein (1972), T. 15, pp. T38-T39, Net investment abroad. Data includes silver trade after 1913, and excludes gold trade. Data for the current account excluding all gold flows for 1850 to 1869 are calculated by taking the current account data from Imlah (1958), less net exports of gold and silver, plus net exports of silver. Data for 1870 to 1920 are calculated by taking the net investment abroad data from Feinstein (1988), less the change in total gold and silver coin stock (as calculated below using data from Capie and Weber (1985)), plus net exports of silver. Data for 1921 to 1944 are taken from Feinstein (1972), net investment abroad, which includes silver trade and excludes gold trade.

Gold:

Gold and silver bullion and specie trade data from 1850 to 1870 are from Imlah (1958), T. 4, pp. 70-72. Gold exports data are from United Kingdom, Board of Trade (various years), exports of gold bullion and specie less imports of gold bullion and specie. Gold exports data for 1917 to 1919 are from Morgan (1952), T. 52, p. 335, net exports of gold coin and bullion. Data on gold in the Bank of England is taken from United Kingdom, Board of Trade (various years), gold in the Bank of England Issue Department, December quarter. Silver and gold imports data were not collected prior to 1858. Data for silver imports from 1850 to 1858 are derived by taking the average ratio of silver imports to total silver trade (export+imports) for the period 1858 to 1945. Data for gold imports from 1850 to 1858 are derived by taking sum of gold and silver exports less estimated silver imports less the data for total gold and silver trade in Imlah (1958), T. 4, pp. 70-72. Silver exports data from Board of Trade (various years), exports of silver bullion and specie less imports of silver bullion and specie. Silver exports data for 1917 to 1919 from Morgan (1952), T. 53, p. 341, net exports of gold and silver, less net exports of gold from T. 52, p. 335.

Data on the change in the monetary stock of gold and silver are from Capie and Weber (1985, T. 7.3, pp. 198-200). Mid year data from 1868 to 1904 were converted to year end data by adding data from following year and dividing by two (this procedure assumes an even distribution across year). Data on the gold coin stock from 1915 to 1920 are calculated by taking the average of the gold coin stock to the total coin stock for the years 1868 to 1914 and multiplying by the figure for the total coin stock. The silver coin stock from 1905 to 1920 is calculated by taking the average of the silver coin stock to the total coin stock for the years 1868 to 1905 and multiplying by the figure for the total coin stock. Data on the change in the monetary gold stock from 1921 to 1931 are taken from the Statistical Abstract (1939), December quarter averages of changes in the gold coin and bullion in the Issue Department of the Bank of England. Data from 1932 to 1945 are from United Kingdom, Financial Secretary of the Treasury (1951), gold reserves in the Bank of England Issue Department and the Exchange Equalisation Account until 1939, then gold and dollar reserves from 1940 to 1945. assumes that the only source of changes in the monetary gold stock arise from changes in the gold holdings of the central bank. This procedure was adopted because of the lack of available data on holdings of gold coin outside the Bank of England.

United States

GDP:

Data for 1869 to 1888 from Kuznets (1961), T. R-25, pp. 561-562, col. 3, GNP, Variant III, 5 year centred moving average and T. R-23, pp. 557-558, col. 1, B. Variant III, billions of dollars. Using annual data from 1889-1892, the 5 year moving average number for 1890 is used to deduce the level of GNP in 1888 as five times the 5 year centred moving average for 1890 less the actual levels for 1889-1892. Using a similar procedure, the annual data for 1869-1888 are derived by disaggregating the 5 year moving averages. Data for 1889 to 1928 are from Kendrick (1961), T. A-IIb, col. 11, pp. 296-297, Gross National Product, Commerce concept, millions of dollars. Data from 1929 to 1945 are from United States, Bureau of the Census (1975), Series F47, Gross National Product, current prices, p. 229. GDP data are calculated by subtracting net income from investments abroad and, net unilateral transfers from abroad from the figures for GNP. The data on income on investments abroad are from United States, Bureau of the Census (1975, series U13, p. 865 then series U5, U6, and U13, p. 864), income on investments abroad, private and public (U5 + U6) less income on foreign investments in US (U13). The data on unilateral transfers are from United States, Bureau of the Census (1975, series U16 and U17, pp. 866-867), unilateral transfers, net private (U16) plus unilateral transfers, net public (U17).

Capital Formation:

Data for 1869 to 1888 from Kuznets (1961, T. R-29, pp. 572-574, col. 1), gross capital formation, five year centred moving average, less col. 3 from T. R-34, pp. 599-600, net changes in claims against foreign countries, five year centred moving average, and T. R-23, pp. 558, col. 4, gross capital formation, B. Variant III, five year centred moving average, less T. R-4, col. 4, net changes in claims against foreign countries. Annual data are derived from the 5 year centred moving average using the same procedure that was applied to the GNP series. Data from 1889 to 1928 are from Kendrick (1961, T. AIIb, col. 7, pp. 296-297), gross private domestic investment, Commerce basis, millions of dollars. Data from 1929 to 1945 are from United States, Bureau of the Census (1975, Series F52, p. 229), gross private domestic investment, current prices. Data on stocks for 1869 to 1888 from Kuznets (1961, T. R-34, pp. 599-600, col. 1), net changes in inventories, current prices, five year centred moving average, and T. R-4, p. 490, col. 3, net changes in inventories, billions of dollars. Annual data are derived from the 5 year centred moving average using the same procedure that was applied to the GNP series, except the data are unscrambled from 1920 backwards. Data from 1889 to 1928 are from Kendrick (1961, T. AIIb, col. 8, pp. 296-297), change in business inventories, millions of dollars. Data from 1929 to 1945 are from United States, Bureau of the Census (1975, Series F60, p. 230), total net change in business inventories, current prices.

Current Account:

Data for 1869 to 1945 from United States, Bureau of the Census (1975), balance on goods and services, series U15, pp. 866-868, plus series U16 and U17, net private and government unilateral transfers. Data from 1869 to 1899 are for fiscal year ended June. Data are converted to calendar year basis by adding each year to subsequent year and dividing by two. For instance, fiscal year data for 1869 are added to fiscal year 1870 and divided by two, yielding calendar year 1869. Data from 1900 are for calendar year. Data to 1873 include exports and imports of gold. Data from 1874-1945 include nonmonetary gold exports. Data for current account less gold for 1869 to 1973 are calculated by taking the current account balance less net exports of gold. Data for current account less gold for 1874 to 1945 are calculated by taking the current account balance less net exports of gold less change in the monetary gold stock (equivalent to subtracting nonmonetary gold exports).

Gold:

Exports: Data for 1869 to 1914 are from NBER Macrohistory Database, series 14112 net gold exports, thousands of dollars, monthly data. Annual data are derived by adding the sum of monthly net exports for each year (fiscal or calendar). Data from 1915 to 1945 from United States, Bureau of the Census(1975), series U197 less U198, gold exports less imports, pp. 884-885, calendar year, millions of dollars.

Monetary gold stock data for 1869 to 1878 are from United States, Bureau of the Census (1975, series X417, p. 993), billions of dollars, annual average. Data from 1879 to 1945 from NBER Macrohistory Database, series 14076, billions of dollars. Data for 1879 to July 1917 are end of month data, then monthly averages of daily figures. Data for fiscal years are end of June data. Data for calendar years are end of December data. Data excludes $287 million in gold coin outside Treasury and Federal Reserve Banks. Increase in value of gold stock from January 1934 to February 1934 of $2.81 billion due to gold revaluation is excluded from figures for changes in monetary gold stock.

1. Net national product if GDP is unavailable.

2. See the notes on capital formation in Australia for a discussion of this point.