¹ Madison (Wise.) Argus, cited in Hibbard, Benjamin H., A History of the Public Land Policies (New York: Macmillan, 1924), p. 221.Google Scholar

² Journal of Economic History, XVII, No. 1 (Mar. 1957), 1–24Google Scholar. The reader is referred to this article for a resume of much of the literature on the subject.

³ Ibid., 2–7.

⁴ Ibid., 24.

⁵ As Hibbard observed, “It would be an endless task to trace the land sales for any considerable area with a view to determine the extent of speculation” (p. 224). Two University of Mississippi graduate students each devoted his master's thesis to the measurement of speculation in only one county. See Russell, Mattie, “Land Speculation in Tippah County 1836–1861” (unpublished M.A. thesis, 1940)Google Scholar; Chapman, Edwin W., “Land Speculation in Tate County 1836–1861” (unpublished M.A. thesis, 1942)Google Scholar. Three University of Nebraska graduate students likewise devoted their theses to land disposal in single counties. See Caylor, John A., “The Disposition of the Public Domain in Pierce County Nebraska” (unpublished Ph.D. dissertation, 1951)Google Scholar; Evans, Evan E., “An Analytical Study of Land Transfer to Private Ownership in Johnson County Nebraska” (unpublished M.A. thesis, 1950)Google Scholar; Stone, James A., “Disposition of the Public Domain in Wayne County, Nebraska, 1868–1893” (unpublished M.A. thesis, 1952).Google Scholar

⁶ Central Iowa is the focus of this study because alphabetical tabulations of Federal land entry records, recently prepared by counsel for the Sac and Fox and Iowa Indian tribes for an Indian claims case, were made available to me. I am indebted to several Iowa abstracting firms for granting free access to their tract books, which facilitated the tracing of resales in the county deed registers. Erling A. Erickson and Michael D. Green assisted in abstracting resale data.

⁷ Duncan, Charles T. (ed.), An Overland Journey from New York to San Francisco in the Summer of 1859 by Horace Greeley (New York: Alfred A. Knopf, 1964), pp. 52–56Google Scholar. For historians who have adopted this position see Gates, Paul W., “The Role of the Land Speculator in Western Development,” Pennsylvania Magazine of History, LXVI, No. 3 (July 1942), 316Google Scholar; Billington, Ray Allen, “The Origin of the Land Speculator as a Frontier Type,” Agricultural History, XIX, No. 4 (Oct. 1945), 206Google Scholar; LeDuc, Thomas, “History and Appraisal of U.S. Land Policy to 1862,” in Ottoson, Howard W. (ed.), Land Use Policy and Problems in the United States (Lincoln: University of Nebraska Press, 1963), pp. 9–15.Google Scholar

⁸ Other researchers have arbitrarily selected other acreage figures as the minimum. Russell, p. iv, and Chapman, p. ii, set the minimum at two thousand acres, but this appears too large by Iowa land-use standards. The same figure was used by Young, Mary E., Redskins, Ruffleshirts, and Rednecks: Indian Allotments in Alabama and Mississippi, 1830–1860 (Norman: University of Oklahoma Press, 1961), p. 99Google Scholar. Minneman, Paul G., “Large Land Holdings and Their Operation in Twelve Ohio Counties” (unpublished Ph.D. dissertation, Ohio State University, 1929)Google Scholar, used five hundred acres, as did Harry Scheiber, N., “State Policy and the Public Domain: The Ohio Canal Lands,” Journal of Economic History, XXV, No. 1 (Mar. 1965)Google Scholar, Table 3, p. 95. Midwestern students have used both one thousand acres and 640 acres. See Caylor, pp. 34–35, and Stone, pp. 44–47.

⁹ The map was adapted from Map Plate CXXXI in U.S. Congress, House, Indian Land Cessions of the United States, by Charles C. Royce (Bureau of American Ethnology, “18th Annual Report, 1896–1897”), 56th Cong., 1st Sess., 1899, H. R. 736. Cession 262 actually contains the major portion of thirty-four counties, but a courthouse fire destroyed the Decatur County book of original entries.

¹⁰ The nine counties were Appanoose, Benton, Boone, Carroll, Hardin, Madison, Marion, Poweshiek, and Wapello. When abstracting, all types of deeds (warranty, quitclaim, and indenture) were used except the following: those that clearly pertained to town lots, those containing metes and bounds descriptions or lot or block numbers that could not easily be translated into sectional descriptions, those reciting nominal considerations (twenty cents per acre or less), those that are illegible, those with obviously defective dates or descriptions, and those with sale dates prior to the date of original entry (such were usually quitclaims deeding only a color of title). When the deed contained two or more descriptions the total consideration was apportioned on the basis of the acreage for each description. When buyers defaulted and tracts were resold, the latter deed was used. Of the 984 entrants with a thousand acres or more in the thirty-three-county area, 299, or about one third, purchased land in the nine selected counties.

¹¹ Breaks in title due to the failure to record deeds, common especially in the early years, made it impossible to chart about 3 per cent of the initial resales. An additional 10 per cent were dropped because the deeds cited nominal considerations of less than twenty cents per acre.

¹² This smaller sample, it was expected, would be representative of the larger group, but a comparison of resales of the two groups in the period 1845–1860 showed that the ten largest entrants were not quite as successful in their speculations as the other large-scale buyers. In the period after 1860, therefore, when the data rest on the activity of the smaller group only, the rates are probably somewhat lower than if all resales of the large buyers had been included.

¹³ All fractional acres were rounded to the nearest whole acre. The entry and sale prices (recorded in dollars and cents) were those given in the books of original entry and deed registers, respectively. The entry price was set at $1.25 per acre if not given in the tract book. All dates were rounded to the nearest whole month. Also included on the data card was a tract identification number corresponding to a tract in the book of original entry, a county and buyer identification number, and a consecutive-data card number. The ten largest original entrants were distinguished from the other large entrymen. Warrant entries were distinguished from cash entries, and tracts sold above $1.25 per acre were differentiated from those sold from twenty cents to $1.25 per acre.

Speculators often entered land for other buyers under a “time entry” agreement (see following). In such cases, borrowers paid land office fees and taxes; agent fees were deducted when loans were repaid. To distinguish credit entries from cash entries (when investors entered land on their own account and paid all fees and taxes themselves), the following rule of thumb was used: credit entries were classified as those in which the land was sold within two years after entry for less than $3.00 per acre or within one year for less than $2.00 per acre. Cash entries were all those in which land was sold twenty-five months or more after entry, or was sold within two years at a greater per acre price than $3.00 or within one year at a greater per acre price than $2.00. Two and three dollars were chosen as safe limits because, at the standard interest rate of 40 per cent on credit sales, approximately 50 cents per year was added to the usual land office price of $1.25 per acre.

Fees and commissions were quite standard for the period. On land warrant entries, agent fees were $5.00 for locating and entering forty-acre tracts, $7.00 for eighties, $8.00 for 120's, and $10.00 for 160's. On cash entries, the fee was reduced by $1.00 for each forty acres since the land office fee was abrogated. Beginning in 1856 when the flood of warrants issued under the liberal bounty act of 1855 (U.S. Statutes at Large, X, 701–2) hit the market, standard agent fees were reduced for warrant entries to $2.50 for forty acres, $5.00 for eighties, $7.50 for 120's, and $10.00 for 160's with the corresponding reduction for cash entries. See advertisements in Iowa City Iowa Capital Reporter, Jan. 11, 1854; Iowa City Daily Evening Reporter, May 28, 1856; Fairfield Iowa Sentinel, Jan. 8, 1857; Fairfield Ledger, Apr. 30, 1857. See also LeGrand Byington, “Circular No. 4, Iowa General Land Agency, Iowa City,” Iowa State Department of History and Archives, Des Moines. Selling commissions varied between 2.5 and 5 per cent of the sale price, but the larger percentage seemed to be more common and was used in this study.

¹⁴ The average yearly tax was based on a “visual mean average” of the annual tax assessments on unimproved lands recorded in the county tax registers. Tax records in three counties had been destroyed; in these cases, the rates of the nearest sample county for which the data were available were substituted—Madison for Appanoose County, Hardin for Boone County, and Marion for Wapello County. Average yearly tax rates rose steadily from one cent per acre in the early years to approximately twenty cents per acre by the 1880's. For a tax series on one section of improved farm land in Muscatine County in eastern Iowa, see Bogue, Allan G., From Prairie to Corn Belt: Farming on the Illinois and Iowa Prairies in the Nineteenth Century (Chicago: University of Chicago Press, 1963), p. 189.Google Scholar

¹⁵ Additional costs might be postage, fees for notarizing deeds and paying taxes, revenue stamps on deeds (required after 1863), title abstract investigations, blank deeds and record books, redemption of tax-deeded land, defalcations of funds by agents, court litigation costs, and faulty assignments of warrants. The cost of investment capital might also be considered as a possible cost item, although scholars disagree sharply on this point. Offsetting factors were that much of the land was entered with warrants purchased at a discount and that some counties allowed payment of the county portion of the taxes with depreciated “county orders.”

Rental income, a much overemphasized aspect of speculation in Congress land in the Midwest, was not a factor in this study. Available evidence indicates that, at least in frontier Iowa, the large speculators primarily engaged in the “time entry,” not the rental, business. For example, over 85 per cent of the 328,508 acres of Congress land in Iowa entered by James S. Easley of Halifax, Virginia—the largest nonresident investor in the Hawkeye State—were secured for second parties under credit arrangements (see Robert P. Swierenga, “Pioneers and Profits: Land Speculation on the Iowa Frontier” [unpublished Ph.D. dissertation, University of Iowa, 1965], ch. vii, tables 1 and 2). Moreover, during the twenty-seven years (1852–1879) of his career in the Iowa land business, the Virginian rented not a single acre of the 47,546 acres which he entered “for future speculation.” As late as the 1870's he insisted that his raw land remain unencumbered with any rental agreements. Even the seemingly innocuous request in 1875 by an Iowa farmer to fence one of Easley's longvacant tracts in exchange for its use as a pasture met with a sharp rejection (James S. Easley to B. F. Yelter, Mar. 20, 1875, Easley Letter Books, XIV, 47, James S. Easley Papers, Alderman Library, University of Virginia, Charlottesville). Easley's non-rental policy apparently was not unique. The extant correspondence of John A. Roebling of Trenton, N.J., Nathaniel Gordon of Exeter, N.H., and Andrew J. Sterrett of Erie, Pa.—all large investors in frontier Iowa in the pre-Civil War decades-similarly makes no reference to rental of agricultural lands. Neither do any of the countless newspaper advertisements of Hawkeye land agents. This is not to assert, however, that all large investors discouraged tenancy. Some may have rented part of their holdings, especially in the postwar period. For example, Paul W. Gates, who several decades ago used the now-unavailable account books and business correspondence of Miles and Elias A. White of Baltimore—second only to Easley, among non-Iowans, in the volume of their Iowa land entries—reported that the Whites “liked to keep tenants on their land” (“The Homestead Act in Iowa,” Agricultural History, XXXVIII, No. 2 [Apr. 1964], 68).Google Scholar

¹⁶ Compound rather than simple interest was used because all contemporary evidence indicates that this was the customary practice. Computing with simple interest, of course, is the standard procedure in many present-day financial arrangements. This method, however, would result in a considerably higher rate of return than the rates here presented because interest is not computed on the accrued interest of previous years. For example, $80 invested for 5 years at 5 per cent simple interest would yield $100, whereas $80 invested for the same period at 4½ per cent compounded annually would likewise yield $100.

Use of constant dollars, often desired by purists in economic studies, was not considered necessary in this study. Unsophisticated mid-nineteenth century investors seemingly were not aware of the subtleties of price comparisons over time. Alternative investments, moreover, offered no escape, since price changes affected all forms of investment, not only capital sunk in land. Converting to constant dollars, in any case, would not have altered significantly the overall average rates of return. In the first place, adding annual taxes and realtor sale commissions to the initial investment provides a built-in hedge against the effect of price changes. During an inflationary period, for example, converting taxes and sale commissions into constant dollars would result in smaller dollar amounts, thus increasing returns; at the same time, converting revenue from land sales into constant dollars would, of course, reduce rates of return. Secondly, general price indexes during the period of this study (1846–1889) fluctuate over time and, therefore, inflationary losses of one period might well be canceled out by gains from subsequent deflationary periods or vise versa. In the pre-Civil War period (1846–1860), the Snyder-Tucker general price index (Base: 1913 = 100) shows a general inflationary trend—from a low of 66.5 in 1846 to a high of 79.6 in 1857—while in the postwar years there is a marked deflation—from a high of 129 in 1864 to a low of 76 in 1886 (see U.S. Bureau of the Census, Historical Statistics of the United States, 1789–1945: A Supplement to the Statistical Abstract of the United States [Washington: U.S. Government Printing Office, 1949], Series LI, pp. 231–32).

¹⁷ See Appendix I for a more complete explanation of the computation procedure. The polynomial equation used to determine interest rate r was Poly = PV (1 + r)^nyear (1 + r X mons/12) + (1 + r X mons/12) σ T(I)(1 + r)^ntyr–1…n — SV = O when PV is the land office price (plus selecting and entering fees), nyear is the number of whole years between the entry and sale date, mons is the fractional months remaining, T(I) is the annual tax, ntyr is the number of whole tax years, and SV is the sale price (less the sale commission). The first factor in the equation compounds interest annually on the initial investment, while the second factor compounds interest annually on taxes. When converting to annual rates of return, all investment periods of less than one full year were computed as being equivalent to one year. This was necessary to avoid gross distortion, although it reduced the rate of return considerably. For example, if an investment of one month's duration yielded a return of 20 per cent, this, rather than 240 per cent (20 X 12), was called the annual rate.

Taxes, computed only on noncredit entries, were determined by multiplying the tract acreage by the tax rate per acre for the respective years. Interest on these yearly tax investments was compounded annually (for a decreasing number of periods) from the date of the first tax payment, due under Iowa law on the second January after entry (see State of Iowa, Revised Code, 1860, pp. 110, 117), until the last January before the sale date. Interest on this final tax payment was computed at simple interest for the period from January 1 of the year of sale until the sale date. The prorated tax amount levied against the property during the calendar year of sale was simply subbuyer. For a column-by-column explanation of the information recorded on the eighty-column data cards and a complete listing of the PROFIT program in FORTRAN IV language for use on an IBM 7040/7044 digital computer, see Swierenga, “Pioneers and Profits,” appendices XV and XVI.

¹⁸ Bogue, and Bogue, , Journal of Economic History, XVII (1957), 22.Google Scholar

¹⁹ These means were obtained by multiplying the entry price of each tract by the rate of return which it earned and dividing the sum of the products by the sum of the entry prices. In addition to listing all of the input data, the computer “print-out” contained the following output data for each tract: the per acre sale price, the absolute percentage increase or decrease between the net entry and net sale price, the gross and net rates of return, and the total time held (in months). For any aggregate group of tracts (either all tracts within one county or all bought or sold in the same year, for example) totals were also printed, listing the total acres in the series, the total entry price, total fees, total sale price, total commissions, weighted average sale price per acre, weighted average gross and net rates of return, weighted average absolute percentage increase or decrease between the net entry and net sale prices, and the weighted average months per acre that all the tracts in the group were held between the entry and sale dates.

²⁰ For figures 2 and 3 I am indebted to Raleigh Barlowe of Michigan State University.

²¹ Specifically, of the 984 large buyers, deed records reveal that 648 (65.9 per cent) lived in Iowa while 310 (31.5 per cent) resided in twenty-four other states and territories plus the District of Columbia. The residence of 26 buyers (2.6 per cent) could not be determined. Only nine states were not represented, seven of these being in the deep South. Biographical data uncovered for 263 Iowans in county histories and manuscript Federal population census rolls indicated that 106, or 40 per cent, were primarily engaged as realtors and bankers during the period of their land entries. Many of these, of course, doubled as lawyers, government officials, merchants, farmers, and even clergymen. The 70 local farmers who comprised the second largest occupational group speculated on a comparatively smaller scale—all but nine entered less than 2,000 acres. Most of the remaining investors were principally employed as merchants, lawyers, physicians, and county officials, with a scattering of land office personnel, manufacturers, carpenters, clergymen, grocers, clerks, and newspaper editors. Information obtained on 81 non-Iowans showed that 25, or almost one third, were realtors and bankers, while another 25 were merchants. The others were mostly farmers and stock-raisers, lawyers, physicians, and government officials.

²² Hibbard (cited in n. 1), pp. 224–25. Cf. p. 221.

²³ Norman Densmore to Benjamin Densmore, Oct. 25, 1855 (Benjamin Densmore Letters, Wisconsin State Historical Society, Madison).

²⁴ U.S. Congress, Congressional Globe, 25th Cong., 2d Sess., 1838, Appendix, p. 547.

²⁵ Although not discussed here, the gross rate of return was also calculated for each tract by ignoring taxes, fees, and commissions. This raised returns by an average of approximately 10 per cent. Readers familiar with the Bogues' study and the generally lower rates of return earned on frontier land investments which they presented will recall the major differences between that study and this, which make comparison difficult. Their data reflected the experience of six specific investor groups, of which the three largest-accounting altogether for over 90 per cent of the acreage in the sample-were largely real estate developers who put tenants on their land. The entries of these investors, moreover, varied widely in time—three acquired their land in the 1830's, two in the 1840's and 1850's, and one in the 1870's. In addition, these investors purchased almost 75 per cent of their sample acreage in the post-Civil War period, 30 per cent of it from land-grant railroads and state governments at prices from two to five times greater than the Congress minimum of $1.25 per acre. Finally, the Bogues lacked precise information on the market price quotations of land warrants.

²⁶ Van Der Zee, Jacob, The Hollanders of Iowa (Iowa City: State Historical Society of Iowa, 1912), p. 71.Google Scholar

²⁷ In six counties in the sample area, 173,064 acres out of 291,378 were entered with warrants. Boone, Hardin, and Madison county tracts were deleted from this part of the analysis because the records do not differentiate between cash and landwarrant entries. In the state as a whole, warrants were used in lieu of cash in 52 per cent of all entries of Congress land.

²⁸ For example, if a settler provided three fifths of the land office price but borrowed the remainder, the average consideration recorded at the time of sale might range below $1.00 per acre, even if the interest cost was included.

²⁹ Easley and Willingham to Head and Russell, Aug. 23, 1867 (Easley Letter Books, VII, 146–47).

³⁰ For the purposes of this study, the midpoint between the monthly high and low retail price quotations, based on 160-acre warrants on the New York and Washington markets, was designated as the “average per acre selling price” for each month. The table of monthly warrant-selling prices in New York and Washington is in my “Pioneers and Profits,” p. 211. The data for the table were kindly supplied by Miss Natalie Disbrow, St. Petersburg, Florida, who is researching the land-warrant market for her doctoral thesis at Cornell University, Ithaca, New York. The title of her study is “Mexican War Bounty Land Warrants.” See U.S. Department of the Interior, Bureau of Land Management, Public Lands Bibliography (Washington: U.S. Government Printing Office, 1962), p. 95Google Scholar. Adding five cents per acre to approximate western rates was arbitrary. The spread between eastern and western retail prices varied somewhat over time, depending on factors of supply and demand. In late 1852, one Iowa agent (Byington, “Circular No. 4”) offered a premium of 15 per cent (ten to twelve cents per acre) above current New York warrant price quotations; in 1855 another local land agent noted that warrants were worth from four to eight cents per acre more in Iowa than in New York (“Letters of J. W. Denison,” Iowa Journal of History and Politics, XXXI, No. 1 [Jan. 1933], 94).Google Scholar

³¹ This is based on actual land warrant costs. If limited only to tracts sold above $1.25 per acre, the rate was 202.0 per cent.

³² In the six counties in which original entry books distinguish land warrant from cash entries, investors suffered net losses on 34,937 acres, or 11.8 per cent of their acreage. However, on sales above $1.25 per acre only, the net loss figure dips to 5,094 acres, or 1.6 per cent—a remarkably small amount.

³³ Easley to J. D. Blair, Feb. 22, 1872 (Easley Letter Books, XI, 201); Easley to E. S. Hedges, Apr. 17,1872 (ibid., 288); Easley to J. P. Casady, July 21, 1874 (ibid., XIII, 220); Easley to G. W. Grance, May 18, 1875 (ibid., XIV, 165).

³⁴ Easley and Willingham to Josiah Bond, Jan. 12, 1869 (ibid., VIII, 365). This was the only such stock offer that Easley accepted, mainly because he could unload some of his unprofitable Wisconsin lands. See letters of Jan. 28 and Feb. 16, 1869 (ibid., 389, 438). In the sixties Easley invested $10,000 in Danville & Richmond Railroad bonds at 10 per cent and $1,300 in United States bonds. Earlier he had put $2,000 in Virginia coupon bonds. See ibid., VIII, 194; IX, 402, 419, 486; XI, 201.