The authors wish to thank their research assistants Arden Hall and Charles Wolin for their help in preparing this paper. The data employed were collected and collated by Wendy Barnes, Sue Boutin, Deborah Doyle, Sheila Moffett, Barbara Robins, and Lynnae Wolin. Financial support was provided by the National Science Foundation, The Institute of Business and Economic Research, and the Computer Center of the University of California, Berkeley. An earlier draft of this paper appeared as Working Paper Number 9 of The Southern Economic History Project. We wish to thank the numerous people who responded with comments and suggestions. Particularly helpful were those of Peter Temin and Robert Gallman.

¹ Otken, Charles H., The Ills of the South or Related Causes Hostile to the General Prosperity of the Southern People (New York: G. P. Putnam's Sons, 1894), pp. 10–11.Google Scholar

² The following brief caricature of historical treatments of Southern growth is intended neither as a thorough summary of the literature, nor as a straw man to be torn apart. Although the analysis has gone through various revisions, the emphasis has remained virtually unchanged from the early works, such as: Grady, Henry W., “Cotton and its Kingdom,” Harper's New Monthly Magazine, Volume LXIII, Number CCCLXXVII (Oct. 1881) pp. 719–34;Google ScholarHammond, M. B., The Cotton Industry: An Essay in American Economic History; Part I, The Cotton Culture and the Cotton Trade (New York: American Economic Association and the Macmillan Company, 1897);Google Scholar and Otken, The Ills of The South; down to the more recent discussions by: Fred A. Shannon, The Farmer's Last Frontier: Agriculture, 1860–1897 Volume V of The Economic History of The United States (New York: Holt, Rinehart and Winston, 1945);Google ScholarWoodman, Harold, King Cotton and His Retainers (Lexington:University of Kentucky Press, 1967);Google ScholarStampp, Kenneth M., The Era of Reconstruction, 1865–1877 (New York:Alfred A. Knopf, 1965);Google Scholar and Salutos, Theodore, Farmer Movements in the South, 1865–1933 (Berkeley:University of California Pess, 1960).Google Scholar

³ Shannon, The Farmer's Last Frontier, pp. 92–93.

⁴ Between 1855 and 1860 the eleven states which later formed the Confederacy appear to have had less than 15 percent of the banks in the United States. For 1855, the Secretary of the Treasury reported 184 Southern banks; for 1860 the figure reported was 198. The five Cotton States (South Carolina, Georgia, Alabama, Mississippi and Louisiana) accounted for about one-third of this total (68 in 1855; 69 in 1860). United States Secretary of the Treasury, Annual Report on the Conditions of Banks in the United States, 1863 38th Congress, 1st Session, House of Representatives, Executive Document Number 3 (Washington: G.P.O., 1863), Tables 10 and 12 pp. 210–14.

⁵ The estimates were based on data in ibid., Tables 10 and 12, pp. 210–14. Because reporting by banks was voluntary, the estimates will only approximate the “average” bank in any region. Our convention throughout this article is to define “the South” as the eleven states which formed the Confederate States of America. Our other regions are defined explicitly in the notes to Table 1, below. However, West Virginia could not be separated from Virginia in the antebellum banking statistics.

⁶ Southern banks did rely more on note issue than deposit creation in comparison with banks in the Northeast. The ratio of bank notes issued to deposits held in 1860 was higher for reporting banks in the South (1.21) than in the twelve Eastern states (0.52). (United States Secretary of the Treasury, Annual Report on the Conditions of Banks … 1863, pp. 210–14).

⁷ While comprehensive data on the banking system of the United States between the organization of the National Banking System in 1863 and the turn of the century is not available, it is nevertheless possible to develop a reasonably accurate picture based upon the Annual Reports of the Comptroller of the Currency. The Comptroller published complete balance sheet data for each National bank continuously beginning in 1863. In response to a request from Congress in 1873, the Comptroller thereafter provided summary statistics for non-National banks who reported voluntarily. Unfortunately, the coverage of non-National banks in these summary tables did not become comprehensive until 1896. The deficiency was particularly severe in the coverage of “private banks.”

In order to provide a better estimate of the number of banks, therefore, we have relied upon the address lists published in Homans’ Bankers’ Almanac between 1865 and 1890. On the other hand, we have employed the Comptroller's data to estimate average deposits, circulation, and size of banks in 1880. We feel that these estimates are sufficiently accurate and comprehensive to support our conclusions.

⁸ Table 1 is based upon the more complete data presented in Appendix Table I.

⁹ The average capital per National bank in the South was $175,000; for the nation as a whole it was $314,000. The data are from United States Comptroller of the Currency, Report for 1869, 41st Congress, 2nd Session, House of Representatives, Executive Document Number 3 (Washington: G.P.O., 1869), pp. 558–93. While consistent and comprehensive data are not available for non-National banks, the figure for National banks is surely an upper bound for Southern banking as a whole, since the National Monetary Commission found few private banks with a capitalization exceeding the $50,000 minimum required by the National Banking System and the average size of state banks was well below the average of National banks. See Bamett, George E., State Banks and Trust Companies Since the Passage of the National Banking Act, in Volume VII of the Publications of the National Monetary Commission (Washington:G.P.O., 1911), p. 205.Google Scholar Also see Table 2 below.

¹⁰ See the data presented in Table 2 below. The average bank in the South had a capital stock of about $133,000; for the United States as a whole the average was $312,000.

¹¹ Sanger, George P. (ed.), The Statutes at large, Treaties, and Proclamations, of the United States of America, from December 1863, to December 1865, Vol. XIII (Boston:Little Brown and Company, 1866), pp. 108Google Scholar, 99–118.

¹² See Sylla, Richard, “Federal Policy, Banking Market Structure, and Capital Mobilization in the United States, 1863–1913,” The Journal of Economic History, XXIX, 1969).Google Scholar Sylla presents a persuasive argument that the restrictions of the National Banking Act effectively curtailed entry into Southern banking (pp. 659–65). He also notes that the resulting monopolization of Southern banking produced higher interest rates on loans than might have existed in a competitive market (pp. 667–70). Additional provisions of the National Banking Act regarding reserves led to a diversion of loanable funds from the country banks to banks in one of the nineteen “reserve cities” (p. 666).

¹³ Sanger, Statutes at Large …, Vol. XIII, p. 484.

¹⁴ The estimated circulation of all state notes fell from 143 million dollars in 1865 to twenty million dollars in 1866 and to less than one million dollars by 1879. See United States Secretary of the Treasury, Annual Report of the Secretary of the Treasury, 1928 (Washington: G.P.O., 1929), p. 552.Google Scholar

¹⁵ Lee, Everett T., “Migration Estimates,” in Kuznets, Simon and Swain, Dorothy (eds.), Population Redistribution and Economic Growth in the United States, 1870–1950, Volume I (Philadelphia:American Philosophical Society, 1957)Google Scholar, Table P-4A, p. 349; and Table P-4B, p. 353.

¹⁶ These figures are computed from: U.S. Census Office, Tenth Census, Statistics of the Population of the United States, Vol. I (Washington:G.P.O., 1883)Google Scholar, Table VI, pp. 416–25. A casual inspection of the data on the number of banks and presence of urban centers suggests the importance of cities to banking in the South. Texas and Virginia accounted for just under a quarter of the South's total population in 1880, yet had 39 percent of all Southern banks (203 of 515) and 43 percent of the state and private banks (175 of 413). About 45 percent (11 of 24) of the cities with a population of eight thousand or under and 40 percent (22 of 55) of all urban centers were located in these two states.

¹⁷ For illiteracy estimates in 1870 see: U.S. Census Office, Ninth Census, The Statistics of the United States … (Washington:G.P.O., 1872), pp. 396–97Google Scholar, and U.S. Census Office, Ninth Census, The Vital Statistics of the United States … (Washington:G.P.O., 1872), pp. 560Google Scholar, 662–63. For illiteracy data in 1880, 1890, and 1900 see: U.S. Census Office, Twelfth Census, Census Reports, Vol. II, “Population, Part II” (Washington: G.P.O., 1902), pp. C-CV.Google Scholar

¹⁸ These figures are based on the data published in United States Comptroller of the Currency, Annual Report for 1880, 46th Congress, 3rd Session, House of Representatives, Executive Document Number 3 (Washington: G.P.O., 1881). State and private banks in the West and South are surely only incompletely covered. However, a check with Homan's Bankers’ Almanac has convinced us that the coverage is sufficiently complete to give a reasonable picture of banking.

David Fand, comparing data from the Comptroller with presumably more complete data obtained when a federal tax was levied on bank deposits between 1876 and 1882, found that deposits in state banks were generally understated by about 5 percent in the Comptrollers Report; and private bank deposits were understated by about 10 percent. A comparison of Fand's regional estimates for deposits in 1880 with the data of Table 2 would show that Fand's estimate of state and private bank deposits was greater in the South by about 13.5 percent, while the rest of the country was about 3.5 percent greater. In the interest of maintaining our earlier regional breakdowns rather than employing that used by Fand (who does not provide state totals), we have chosen to employ the Comptroller's data. See Fand, David I., “Estimates of Deposits and Vault Cash in the Non-National Banks in the Post Civil War Period in the United States: 1876–1896,” (Unpublished Ph.D. dissertation, University of Chicago, March 1954).Google Scholar

¹⁹ Davis, Lance, “The Investment Market, 1870–1914: The Evolution of a National Market,” The Journal of Economic History, XXV, (Sept. 1965)Google Scholar employs data from the Comptroller of the Currency and other sources, to show the tendency for regional capital markets to evolve into a national market in the period 1870–1914. He singles out the South as an exception to this trend, particularly in the development of flows of long-term investment capital between regions (pp. 388–93). Richard Sylla, “Federal Banking Policy …,” notes the extent to which the National Banking System encouraged flows of funds out of agricultural regions to the industrial markets.

²⁰ For National banks in 1869 the ratio of loans to all assets was 36 percent in the Cotton States, and reports from twenty-three state banks in Georgia and South Carolina that year give their percentage of loans as 45 percent. (United States Comptroller Report for 1869, pp. CXL-CXLV). This compares to 60 percent or more for banks in the East and West. Of course, the prohibition on mortgages by National banks substantially cut the level of loans on real estate of National banks in all regions.

²¹ United States Comptroller of the Currency, Annual Report for 1890, 51st Congress, 2nd Session, House of Representatives, Executive Document Number 3 (Washington: G.P.O., 1890).Google Scholar Davis, “Investment Market …” explores this and supporting data further to show the regional differentials in mortgage rates. A report of the 1890 Census showed that the eleven Southern states accounted for only 10.7 percent of all farm mortgages, though 33.4 percent of the Nation's farms were in those states. See Holmes, George K. and Lord, John S., United States Census Office, Eleventh Census, Report on Real Estate Mortgages in the United States, Volume XII (Washington, G.P.O., 1895), p. 174.Google Scholar

²² United States Comptroller of the Currency, Annual Report for 1890, p. CXXXIX.

²³ On the flood of Northern merchants, discharged soldiers and other entrepreneurs into the Southern furnishing business, see the remarks of Reid, Whitelaw, After the War: A Southern Tour, May 1, 1865 to May 1, 1866 (Cincinnati and New York: Moore, Wilstach and Baldwin, 1866), pp. 481–82;Google ScholarSomers, Robert, The Southern States Since the War: 1870 (London and New York: Macmillan Company, 1871), pp. 70Google Scholar, 214–43; and Clark, Thomas D., “The Furnishing and Supply System in Southern Agriculture since 1865,” The Journal of Southern History, XII, (Feb. 1946), 22–44.Google Scholar

²⁴ Typical of the comments on the increasing wealth of merchants are the remarks by Otken, Ills of the South, who claimed that men engaged in the furnishing business “… managed to accumulate handsome fortunes, varying from $10,000 to $20,000.” (p. 80), Grady, “Cotton and its Kingdom,” and Harry Hammond in a report for the South Carolina Board of Agriculture, South Carolina: Resources and Population, Institutions and Industry (Charleston: Walker, Evans and Cogswell, 1883)Google Scholar, make similar references to the prosperity of storekeepers.

²⁵ Hammond, The Cotton Industry, p. 153, reported from surveys by the Georgia Department of Agriculture over the period 1880 to 1890 and by the Louisiana Commissioner of Agriculture between 1886 and 1896. A survey by Jones, W. N., First Annual Report of the Bureau of Labor Statistics of the State of North Carolina, for the Year 1887 (Raleigh:J. Daniels, 1887)Google Scholar produced similar results. Evidence abounds that these conditions existed throughout the South. See Otken, The Ils of the South, ch. xi, and Hilgard, Eugene W., editor, United States Census Office, Tenth Census, Report on Cotton Production in the United States …, Two Parts (Washington, G.P.O., 1884).Google Scholar

²⁶ A total of 71,000 farms and 2,909 stores were included in the cotton regions of South Carolina. (Hammond, South Carolina …).

²⁷ The figure of one hundred dollars is based on an analysis of data from the 1880 manuscript Census of Agriculture.

²⁸ On the early passage of lien laws see: Zeichner, Oscar, “The Legal Status of the Agricultural Laborer in the South,” Political Science Quarterly, LV, (Sept. 1940), 412–28;CrossRefGoogle ScholarBanks, Enoch M., “The Economics of Land Tenure in Georgia,” Studies in History, Economics, and Public Law, Vol. XXIII, No. 1 (1905);Google Scholar and Brooks, Robert P., “The Agrarian Revolution in Georgia, 1865–1912,” Bulletin of the University of Wisconsin, History Series, III (1914).Google Scholar There seems little doubt that the new system was general throughout the South. See the references listed in fns. 2 and 25 above.

²⁹ In addition to Shannon, Harold Woodman, King Cotton …, has argued that acquisition of land by merchants was an increasing tendency throughout the latter half of the nineteenth century in the South (p. 311). The extensive study of tenant plantations in the 1910 Census indicated that the extent of multi-farm ownership in the South was considerable. Such a phenomenon would be consistent with but not proof of the rise of a merchant landlord. United States Bureau of the Census, “Plantations in the South,” ch. xv, Agriculture, … Volume V of the Thirteenth Census …, 1910 (Washington, G.P.O., 1913), pp. 877–93.Google Scholar

³⁰ Under slavery, blacks were compelled to work and could not consider any of their time as their own. Once free, they chose to consume a portion of their time in leisure. Women and children, who before the War were used as labor in the fields, opted to remain at home. The men who offered their labor for wages also expressed a preference for leisure which exhibited itself in an unwillingness to work on Sundays and occasionally Saturdays as well, along with a desire for shorter hours than was customary under slavery. These changes produced a marked upward shift in the supply curve of labor which commentators estimated to account for a decline of as high as one-third in the quantity of Negro labor supplied. See the contemporary comments reported by Loring, F. W. and Atkinson, C. F., Cotton Culture and the South Considered with Special Reference to Emigration (Boston:A. Williams and Company, 1869)Google Scholar, particularly pp. 8–9, 13, 15, 20, 22–23, and 110.

³¹ This argument can be viewed as supportive of Evsey Domar's hypothesis that if land is “free” and labor is relatively scarce, then no rent can be extracted from land. This will encourage restrictions on labor mobility. We suggest that the Southern landlord might have allowed the merchant to bind the tenant to the farm, thus assuring a labor supply. He is then able to extract “rent” from the labor due to its immobility. Domar, Evsey D., “The Causes of Slavery or Serfdom: A Hypothesis,” The Journal of Economic History, XXX, (March 1970), 18–32.CrossRefGoogle Scholar

³² For example, see the crop-lien contract reproduced in “Southerner,” “Agricultural Labor at the South,” Galaxy Magazine, Vol. XII, No. 9 (Sept. 1871), 338.Google Scholar Bull disputes the prevalence of the one-crop lien, and it is true that almost all liens extended to any crop the farmer produced. However, such provisions served merely to give added security to the debt in the case of a failure of the cotton crop. Bull, Jacqueline P., “The General Merchant in the Economic History of the New South,” Journal of Southern History, XVIII, (Feb. 1952), 41–42.Google Scholar

³³ The quotations are from Jones, Report of the Bureau of Labor Statistics of … North Carolina …, 1887, pp. 88–89, p. 92, and p. 129 respectively. For similar contemporary opinions see Otken, Ills of the South, pp. 54–64; and Smith, Eugene A., “Cotton Production in Alabama,” in Hilgard, Report on Cotton Production, Vol. II, pp. 62–63, 156.Google Scholar Also see Hammond, , The Cotton Industry, Vol. I, pp. 150–52.Google Scholar Frequent testimony before the United States Senate Committee on Agriculture and Forestry in 1895 is evidence that these practices of the merchants continued at least into the mid-nineties. “Present Condition of Cotton-Growers of the United States’ Compared with Previous Years,” Report of the Committee on Agriculture and Forestry, February 23,1895, 53rd Congress, 3rd Session, Senate Report Number 986, Part I (Washington, G.P.O., 1895).Google Scholar

³⁴ The classic statement of this position can be found in Grady: “The first reform, however, that must be made is in the system of fanning. The South must prepare to raise her own provisions, compost her own fertilizers, cure her own hay, and breed her own stock. Leaving credit and usury out of the question, no man can pay seventy-five cents a bushel for corn, thirty dollars a ton for hay, twenty dollars a barrel for pork, sixty cents for oats, and raise cotton for eight cents a pound.” “Cotton and its Kingdom,” p. 723. Also see Jones, Bureau of Labor Statistics of … North Carolina … 1887, pp. 76–77.

³⁵ Robert Galhnan has argued that antebellum Southern agriculture was largely self-sufficient in the provision of foodstuffs. “Self-Sufficiency in the Cotton Economy of the Antebellum South,” in Parker, William N. (ed), The Structure of the Cotton Economy of the Antebellum South (Washington:The Agricultural History Society, 1970), pp. 5–23.Google Scholar This conclusion upset the traditional view that slave plantations specialized in cotton production to the exclusion of food crops.

³⁶ This decline cannot be explained by a shift toward production of other foodstuffs. While the Census is incomplete on the production of miscellaneous crops, the published data indicate a decline in the per capita production of wheat, oats, barley, rice, buckwheat, and rye. This decline in per capita production represented more than just a failure of Southern agriculture to keep pace with the population growth. The same decline in food production is noted when computed on a rural population base.

³⁷ Hammond, The Cotton Industry, Vol. I, p. 153.

³⁸ Hilgard, Report on Cotton Production …, Two Parts.

³⁹ It was decided not to attempt an extensive analysis of the 1870 returns for two reasons. First, that Census severely under-enumerated the Southern population, particularly Negroes. A second disadvantage from our point of view was the failure of the Agricultural Census to record the tenure of the farm operator.

⁴⁰ The manuscript reports of population have been retained by the United States National Archives. For details see Davidson, Katherine H. and Ashby, Charlotte M., Compilers, “Records of the Bureau of the Census,” The National Archives, Preliminary Inventories, Number 161 (Washington, G.P.O., 1964), p. 101.Google Scholar The manuscript schedules for agriculture for fifteen Southern states are available from the University of North Carolina Library. For details on this collection see Boone, Samuel M., “Agricultural and Manufacturing Census Records of Fifteen Southern States for the Years 1850, 1860, 1870, and 1880” (Chapel Hill:University of North Carolina Library, 1966).Google Scholar

⁴¹ For a discussion of the sampling procedure see Ransom, Roger, Sutch, Richard, and Boutin, George, “A Sample of Southern Farms in 1880: Sampling Procedure,” Southern Economic History Project, Working Paper Series, Number 2 (Berkeley:Institute of Business and Economic Research, University of California, September 1969)Google Scholar, and Ransom, Roger and Sutch, Richard, “Economic Regions of the South in 1880,” Southern Economic History Project, Working Paper Series, Number 3 (Berkeley:Institute of Business and Economic Research, University of California, March 1971).Google Scholar

⁴² Gallman, “Self-sufficiency ….”

⁴³ Rather than employ the figures used by Gallman based on the caloric content of various crops, we have used conversion ratios based on the total digestible nutrients available to the animal as reported in Henry, W. A. and Morrison, F. B., Feeds and Feeding (17th ed., Madison:Henry-Morrison Company), p. 121.Google Scholar For the case of sweet potatoes, we have used the data in Morrison, F. B., Feeds and Feeding (20th ed., Ithaca:Morrison Company, 1940), p. 997.Google Scholar The conversion ratios computed from these sources are: barley, 0.866; buckwheat, 0.866; oats, 0.433; rye, 1.05; wheat, 1.104; cowpeas, .946; dried beans, .946; Irish potatoes, 0.220; and sweet potatoes, 0.362.

⁴⁴ Seed allotments can vary considerably. There is a wide range where substitution of labor inputs for seed inputs is possible particularly with corn, peas, beans, and potatoes. The practice in the South in 1880 was apparently based on well-known rules of thumb which tended to be seed intensive. Records of actual practices, however, are not abundant and those which are reported in the agricultural journals and experimental station reports are almost always exceptional cases. Rather than use these reports, we have relied on the recommended practice reported in Robinson, Solon, Facts for Farmers, 2 Vol. (New York:A. J. Johnson, 1867)CrossRefGoogle Scholar, and Bailey, L. H., Farm and Garden Rule Book (18th ed.; New York:Macmillan Company, 1912).Google Scholar These books recommended labor intensive cultivation practices which produced higher yields from a given quantity of seed than was likely to be common in the South in 1880. Thus our estimates will tend to exaggerate the net grain output. The percentages deducted for seed in the calculations are as follows: com, 2 percent; wheat, rye, and barley, 9 percent; oats and buckwheat, 7 percent; peas and beans, 8 percent; Irish and sweet potatoes, 3 percent. See Bailey, pp. 92–93, and Robinson, pp. 673–74, 716, 750, 690–92, and 699. Also see Seaman, Ezra C., Essay on the Progress of Nations (New York:Scribners, 1852), pp. 275Google Scholar, 453, and 625; Gallman, “Self-Sufficiency …, ” and Raymond C. Battalio and John Kagel, “The Structure of Antebellum Southern Agriculture: South Carolina, A Case Study,” in Parker, The Structure of the Cotton Economy, p. 28.

⁴⁵ Gallman relied primarily on the report by the United States Department of Agriculture, Consumption of Feed by Livestock, 1909–1956, Production Research Report 21 (Washington: G.P.O., 1958).Google Scholar Also see Battalio and Kagel, “The Structure of Antebellum Southern Agriculture” and Hutchinson, William K. and Williamson, Samuel H., “The Self-Sufficiency of the Antebellum South: Estimates of the Food Supply,” Bureau of Business and Economic Research, Working Paper Series, Number 71–6 (Iowa City, College of Business Administration:The University of Iowa, March 1971).Google Scholar

⁴⁶ Among many references, the most useful we found was Dancy, F. B., “Stock Feeding, as Practised in North Carolina,” North Carolina Experiment Station Bulletin, No. 66 (September 15, 1889)Google Scholar, and Georgia State Department of Agriculture, A Manual on the Hog, Circular No. 40 (Atlanta: James P. Harrison and Company, June 15 1877).Google Scholar

⁴⁷ Henry and Morrison, Feeds and Feeding.

⁴⁸ United States Department of Agriculture, Bureau of Agricultural Economics, Livestock on Farms, January 1, By States: Revised Estimates of Numbers, Value per Head, and Total Value, 1867–1919 (Washington: G.P.O., January 1938).Google Scholar

⁴⁹ See Appendix Table II for the definition of each region, the sampled counties from which the data were drawn, and the number of farms included in the sample.

⁵⁰ The energy value of ten bushels of corn would be approximately 450,000 calories. Henry and Morrison, Feeds and Feeding, p. 998. This implies a daily caloric intake of less than 1250 calories, substantially below the levels required for adults.

⁵¹ Our sample of farms illustrates this fact for 1879. In each of the counties sampled the proportional variance in the physical yields per acre for corn were higher than for cotton. Moreover, throughout the period, farm gate prices of corn fluctuated more sharply than did cotton prices.