The author wishes to thank Lee Alston, Alberta Charney, Price Fishback, Jim Griffin, Robert Higgs, Betsy Hoffman, Mark Isaac, Ron Johnson, Ken Kroner, Ron Oaxaca, Ed Ray, Barbara Sands, Tom Weiss, this JOURNAL's referees, and participants in the 1985 Cliometrics Conference, the 1985 All-California Economic History Conference, and workshops at the University of Kansas, the University of Southern California, and Washington University. Statistical support was provided by Cliff Mangano.Google Scholar

¹ The cartel literature is both varied and large. A sample includes Stocking, George W. and Watkins, Myron W., Cartels in Action (New York, 1946);Google ScholarFog, Bjarke, “How are Cartel Prices Determined?” Journal of Industrial Economics, 3 (11 1956), pp. 16–23;CrossRefGoogle ScholarMacAvoy, Paul W., The Economic Effects of Regulation: The Trunk-Line Railroad Cartels and the Interstate Commerce Commission Before 1900 (Cambridge, MA, 1965);Google ScholarKolko, Gabriel, Railroads and Regulation, 1877–1916 (New York, 1965);CrossRefGoogle ScholarHay, George A. and Kelly, Daniel, “An Empirical Survey of Price Fixing Conspiracies,” Journal of Law and Economics, 17 (04 1974), pp. 13–39;CrossRefGoogle Scholar and Ulen, Thomas, “Railroad Cartels Before 1887: The Effectiveness of Private Enforcement and Collusion,” Research in Economic History, 8 (1983), pp. 125–44.Google Scholar There is also a large game-theoretic literature. See Tirole, Jean, The Theory of Industrial Organization (Cambridge, MA, 1988), for a survey.Google Scholar

² Examples include Joskow, Paul, “Inflation and Environmental Concern: Structural Change in the Process of Public Utility Regulation,” Journal of Law and Economics, 17 (10 1974), pp. 291–328;CrossRefGoogle ScholarHallagan, William, “Contracting Problems and the Adoption of Regulatory Cartels,” Economic Inquiry, 23 (01 1985), pp. 37–56;CrossRefGoogle ScholarShepard, Lawrence, “Cartelization of the California- Arizona Orange Industry, 1934–1981,” Journal of Law and Economics, 29 (04 1986), pp. 83–121;CrossRefGoogle ScholarCave, Jonathan and Salant, Stephen W., “Cartels that Vote: Agricultural Marketing Boards and Induced Voting Behavior,” in Bailey, Elizabeth, ed., Public Regulation: New Perspectives in Institutions and Policies (Cambridge, MA, 1987);Google Scholar and Wiggins, Steven N. and Libecap, Gary D., “Firm Heterogeneities and Cartelization Efforts in Domestic Crude Oil,” Journal of Law, Economics and Organization, 3 (Spring 1987), pp. 1–25.Google Scholar

³ For an overview of research and disagreements about OPEC, see Gately, Dermot, “A Ten-Year Retrospective: OPEC and the World Oil Market,” Journal of Economic Literature, 22 (09 1984), pp. 1100–14;Google Scholar and Griffin, James M., “OPEC Behavior: A Test of Alternative Hypotheses,” American Economic Review, 75 (12 1985), pp. 954–63.Google Scholar

⁴ See de Chazeau, Melvin G. and Kahn, Alfred E., Integration and Competition in the Petroleum Industry (New Haven, 1959), pp. 123–25;Google Scholar and Zimmermann, Erich W., Conservation in the Production of Petroleum (New Haven, 1957), pp. 59, 213–24.Google Scholar

⁵ See Adelman, Morris, The World Petroleum Market (Baltimore, 1972), p. 148;Google Scholar and Lovejoy, Wallace F. and Homan, Paul T., Economic Aspects of Oil Conservation Regulation (Baltimore, 1967), pp. 226–27.Google Scholar

⁶ U.S. crude oil production by state is in American Petroleum Institute, Petroleum Facts and Figures (Washington, DC, 1971), pp. 68–69. California had large oil reserves, but until the postwar period it served a separate West Coast market.Google Scholar

⁷ For discussion of common pool conditions in crude oil production, their impact on prices, and efforts to secure government controls, see Libecap, Gary D. and Wiggins, Steven N., “Contractual Responses to the Common Pool: Prorationing of Crude Oil Production,” American Economic Review, 74 (03 1984), pp. 87–98;Google Scholar and Libecap, Gary D., “The Political Allocation of Mineral Rights: A Re-Evaluation of Teapot Dome,” this JOURNAL, 44 (06 1984), pp. 381–91.Google Scholar

⁸ This decline in nominal crude oil prices far exceeds the decline in the consumer price index over the same period. See U.S. Department of Commerce, Bureau of the Census, Historical Statistics of the Uxnited States, Colonial Times to 1970 (Washington, DC, 1975), p. 211.Google Scholar

⁹ Oil production costs are largely a function of geology and time. Newly developed reservoirs provide oil at very low marginal extraction costs, since underground pressures are sufficient to expel the oil to the surface once the reservoir is punctured by a well. Over time those pressures are dissipated, particularly if the field is exploited rapidly, as is likely under common pool conditions. Eventually, installation of pumping wells and the injection of natural gas, water, or other substances into the reservoir become necessary to extract the oil, but these actions raise marginal extraction costs.Google Scholar

¹⁰ The Oklahoma City and East Texas fields were the largest during the 1926 to 1935 period, but East Texas dwarfed all others, being 10 to 12 times as large in area as Oklahoma City. See Oil Weekly, April 24, 1931;Google Scholarand Oil Weekly, February II, 1935.Google Scholar

¹¹ Crude oil reserves data are from the Oil and Gas Journal, January 18, 1937;Google Scholarand the American Petroleum Institute, Reserves of Crude Oil, Natural Gas Liquids, and Natural Gas in the United States and Canada as of December 31, 1979 (Washington, DC, 1980), pp. 25–66. Proved reserves are defined by the American Petroleum Institute as crude oil that is recoverable with existing technology.Google Scholar

¹² Stripper wells are defined in Lovejoy and Homan, Economic Aspects, p. 185.Google Scholar

¹³ Constantin v. Smith, 57 F.2nd 227 (February 1932).Google Scholar

¹⁴ Early efforts for interstate market-demand prorationing are outlined in Oil Weekly, January 23, 1931; April 10, 1933;Google ScholarMurphy, Blakely M., Conservation of Oil and Gas: A Legal History, 1948 (Chicago, 1949), pp. 545–47;Google Scholar and Oil and Gas Journal, March 23 and 30, 1933.Google Scholar

¹⁵ The Bureau of Mines issued a Monthly Petroleum Statement for each state, which summarized current output, oil stocks (surface inventories), and forecasts of next month's demand. These estimates were for total U.S. demand, which was then divided among the states as a form of quota. This information was supplied to each of the state agencies in enough time to set next month's authorized output level. The output data were gathered by the bureau from crude oil producers, the Interstate Oil Compact Commission, the American Petroleum Institute, operators of pipelines and other means of transport, and refineries. The bureau based its forecasts of demand for crude oil on estimates of the demand for gasoline, fuel oil, kerosene and other petroleum products. Its procedures are discussed in the Cole Committee Hearings, Petroleum Investigation: Hearings before a Subcommittee of the Committee on Interstate and Foreign Commerce on H.R. 441, 73rd Congress, 2nd session, part 2 (Washington, DC, 1934), p. 1298;Google Scholar and Libecap, Gary D., “The Political Economy of the Establishment of the Interstate Oil Cartel, 1933–1940,” in Higgs, Robert, ed., Emergence of the Modern Political Economy (Greenwich, CT, 1985), pp. 53–81.Google Scholar

¹⁶ Panama Refining v. Ryan, 293 U.S. 388 (January 1935).Google Scholar

¹⁷ “Connally Hot Oil Act,” U.S. Statutes Ar Large, 49 (February 1935), p. 30.Google Scholar

¹⁸ Discussion of the impact of the NRA on Texas is provided in Wiggins and Libecap, “Firm Heterogeneities,” pp. 18–19.Google Scholar

¹⁹ Article V of the compact stated that the purpose of the agreement was not to fix or stabilize prices. There were two barriers to explicit discussion of price fixing within the compact. One was a fear of violating federal antitrust laws; the other was Texas's concern that formal compact rules to equate interstate output to market-demand estimates would lead to regulation of its production by the Interstate Oil Compact Commission rather than by the Railroad Commission. For discussion of the pressure placed on compact members to conform to output rules, see Oil and Gas Journal, July 27, 1939.Google Scholar

²⁰ U.S. demand and production data are from the U.S. Department of the Interior, Bureau of Mines, Minerals Yearbook (Washington, DC, 1934–1973).Google ScholarAs noted in fn. 15, the bureau provided total and state demand estimates each month to the state regulatory agencies. These demand forecasts for crude oil were based on estimates of demand for major oil products. See Minerals Yearbook, 1937, p. 978. Domestic crude oil demand figures are provided from 1943 through 1972. From 1933 through 1942 the demand estimates are for domestic crude oil and imports. Crude oil imports were small during that period, 2 percent of U.S. crude oil production in 1937, as reported by the American Petroleum Institute, Petroleum Facts and Figures, p. 286. Imports of crude oil and refined products do not become a significant part of the U.S. market until later. In 1955 imports of crude oil were 11 percent of U.S. crude oil production. The bureau also published demand estimates for all oils, which include crude oil, refined products, and natural gas.Google Scholar

²¹ Posted nominal crude oil prices are from American Petroleum Institute, Petroleum Facts and Figures, p. 449; and Oil and Gas Journal. Where there are discrepancies the Oil and Gas Journal price is used. Spot prices generally are not available, but movements in posted prices reflect the general path of spot prices.Google Scholar

²² Reserve estimates are from American Petroleum Institute, Reserves of Crude Oil, p. 25. It is possible that discoveries prior to 1933 represented larger percentage increases in total U.S. reserves than in the later period, contributing to the greater volatility of prices between 1913 and 1933 as shown in Figure 2. Nevertheless, an examination of reserve data, which are available after 1934, reveals large shifts in reserves from new discoveries which, if uncontrolled, could have caused prices to fall sharply. For example, between 1938 and 1939 reserves rose by over 36 percent. The successful efforts by the Interstate Oil Compact to address that problem associated, in part, with new Illinois production are described in the text. Similarly, U.S. reserves rose by nearly 32 percent between 1969 and 1970 with the discovery of the Alaska North Slope fields with no significant impact on prices.Google Scholar

²³ Stocks or inventories are from de Chazeau, and Kahn, , Integration and Competition, pp. 137–66; and American Petroleum Institute, Petroleum Facts and Figures, p. 109.Google Scholar

²⁴ Nominal crude oil prices for 36-degree gravity oil are deflated by the Consumer Price Index for all items,Google Scholar found in U.S. Department of Commerce, Historical Statistics, p. 210.Google Scholar

²⁵ See Joskow, “Inflation and Environmental Concern”, pp. 291–328.Google Scholar

²⁶ See the following issues of Oil and Gas Journal: March 20, 1941; January. 22, 1947; December 28, 1950; August. 25, 1951; September. 20, 1954; April. 14, 1957; June 6, 1960; February. 25, 1963; and January. 11, 1971.Google Scholar

²⁷ It is not the aim of this article to explicitly model cartel behavior among the states. Griffin, “OPEC Behavior,” tests various production hypotheses for OPEC and finds coordination among OPEC members in output patterns. He does not find evidence that Saudi Arabia acts as a residual producer. While he does not examine the reasons for market-sharing behavior, the consistency of the results reported here with his suggest that political conditions may make residual production too politically costly, when governments are the cartelizing parties. Political and production conditions among the states are not stable over the 40-year period examined here. Comparisons of the coefficients of variation across the states avoid making any assumptions regarding the nature and stability of the underlying relationships. Estimations using Griffin's approach were investigated, and they also reveal market-sharing behavior among the states. For the purposes of this article, examination of the coefficient of variations is the preferable approach for determining the extent of coordination among the states.Google Scholar

²⁸ For a discussion of the test for cointegration and error correction, see Engle, Robert F. and Granger, C. W. J., “Cointegration and Error Correction: Representation, Estimation and Testing,” Econometrica, 55 (03. 1987), pp. 251–76.CrossRefGoogle Scholar See also Maddala, G. S., Introduction to Econometrics (New York, 1988), pp. 212–21, for a discussion of estimation techniques. A stationary time series implies that the relationship between the variables does not change significantly over time.Google Scholar

²⁹ The value of minerals production is from the U.S. Department of Commerce, Bureau of the Census, Statistical Abstract of the United States 1939 (Washington, DC, 1940), p. 734. Value of agricultural products is from U.S. Department of Commerce, Bureau of the Census, Census of Agriculture 1935, Statistics for State and Counties, vol. 2, part 3, 2nd series (Washington, DC, 1936), pp. 742–43.Google Scholar

³⁰ See testimony regarding the need to restrain East Texas in the Texas Senate Journal (July 1931), pp. 418–52;Google Scholarand Texas House Journal (July 1931), pp. 428–59.Google Scholar

³¹ See Oil Weekly, July 17, 1931;Google Scholarand 42nd Texas Legislature, Acts, 1931, regular session, chap. 50, art. 6049.Google Scholar

³² Oil Weekly, february 27, 1933;Google Scholar and Prindle, David F., Petroleum Politics and the Texas Railroad Commission (Austin, 1981), p. 50.Google Scholar

³³ Hostility in Texas toward major oil producers has been noted in a variety of studies. See Prindle, Petroleum Politics;Google Scholar and Pratt, Joseph A., “The Petroleum Industry in Transition: Antitrust and the Decline of Monopoly Control in Oil,” this JOURNAL, 60 (12. 1980), pp. 815–37.Google Scholar

³⁴ Lovejoy and Homan, Economic Aspects, p. 114.Google Scholar

³⁵ The number of producing days was a full-time equivalent number. In practice, firms produced for a fraction of each day during the month, calculated from the number of producing days authorized by the Texas Railroad Commission. Capacity production is defined in various ways (see Lovejoy and Homan, Economic Aspects, pp. 97–102). In general, it is maximum production for a specified period from existing wells.Google Scholar

³⁶ See Libecap and Wiggins, “Contractual Responses,” for discussion of the initial prorationing allocation scheme in Texas. For discussion of acreage and depth, see Lovejoy and Homan, Economic Aspects, p. 144.Google Scholar

³⁷ Railroad Commission v. Humble Oil and Refining Co., 193 S.W.2nd 824 (March. 1946).Google Scholar

³⁸ Oil and Gas Journal, May 8, 1961.Google Scholar

³⁹ These shares are state output as a percentage of total prorationed output from California, Illinois, Kansas, Louisiana, Oklahoma, and Texas.Google Scholar

⁴⁰ Oil and Gas Journal, August. 4, September. 9, and October. 6, 1938.Google Scholar

⁴¹ Ibid., August. 24, 31, 1939; October. 6, 1939.

⁴² Murphy, Conservation of Oil and Gas, pp. 92–120, outlines conditions in Illinois. Fear of federal intervention helped to persuade Illinois firms to support stricter state prorationing legislation.Google Scholar

⁴³ Imports also were a concern to the state regulatory agencies. In 1955 imports rose to II percent of U.S. domestic production. The state agencies, led by the Texas Railroad Commission, lobbied Congress for import controls. For discussion of U.S. oil import policies, see Oil and Gas Journal, March. 17, 1958;Google Scholar and Barzel, Yoram and Hall, Christopher D., The Political Economy of the Oil Import Quota (Stanford, 1977).Google Scholar

⁴⁴ Oil and Gas Journal, July 28, 1958.Google Scholar

⁴⁵ Ibid., July 6, 1960; February. 25, 1963.

⁴⁶ Ibid., December 14, 1959; July 20, 1960; December. 20, 1960.

⁴⁷ Ibid., December. 17, 1962.

⁴⁸ Ibid., April 22, 1961.