1. Burnham, Walter Dean, “Party Systems and the Political Process,” in Chambers, William N. and Burnham, Walter Dean, eds., The American Party Systems, 2nd ed. (New York: Oxford University Press, 1975), p. 297Google Scholar.

2. Skowronek, Stephen, Building the New American State (New York: Cambridge University Press, 1982)CrossRefGoogle Scholar.

3. Bensel, Richard, Sectionalism and American Political Development, 1880–1980 (Madison: University of Wisconsin Press, 1984)Google Scholar.

4. Orloff, Ann S. and Skocpol, Theda, “Why Not Equal Protection?” American Sociological Review, 49 (1984): 726–750Google Scholar;Weir, Margaret, Orloff, Ann S., and Skocpol, Theda, The Politics of Social Policy in the United States (Princeton, NJ: Princeton University Press, 1988)Google Scholar; Christopher Howard et al., “Government Institutions, Women's Associations, and the Enactment of Mothers' Pensions in the United States, 1910–1935” (paper presented at the annual meeting of the American Political Science Association, Washington, DC.).

5. Evans, Peter B., Rueschemeyer, Dierich, and Skocpol, Theda, Bringing the State Back In(New York: Cambridge University Press, 1985)CrossRefGoogle Scholar.

6. Arrow, Kenneth, Social Choice and Individual Values (New Haven, CT: Yale University Press, 1951)Google Scholar;McKelvey, Richard D., “Intransitivities in Multidimensional Voting Models and Some Implications for Agenda Control,” Journal of Economic Theory, 16 (1976): 472–482Google Scholar.

7. Weingast, Barry R. and Marshall, William J., “The Industrial Organization of Congress,” Journal of Political Economy, 96 (1988): 132–163Google Scholar;, McNollgast, “Structure and Process. Politics and Policy,” Virginia Law Review, 75 (1989): 431–482Google Scholar.

8. Shepsle, Kenneth and Weingast, Barry, “Structure-Induced Equilibrium and Legislative Choice,” Public Choice, 37 (1981): 503–519Google Scholar; cf. Krehbiel, Keith, “Spatial Models of Legislative Choice,” Legislative Studies Quarterly, 13 (1988): 259–320Google Scholar.

9. Weingast and Marshall, “The Industrial Organization of Congress.”

10. Horn, Murray, “Political Economy of Public Administration” (Ph.D. dissertation, 1990)Google Scholar; McNollgast, “Structure and Process, Politics and Policy”; Moe, “Political Uncertainty.”

11. To our knowledge, there is no systematic, embracing approach to the effect of institutions on elections. Instead, there are several, almost separate, literatures on parts of the question: (1) the issues of redistricting and gerrymandering (Grofman, Bernard et al. , eds., Representation and Redistrictmg Issues [New York: Macmillan, 1982]Google Scholar; Cain, Bruce E., The Reapportionment Puzzle [Berkeley: University of California Press, 1984])Google Scholar; (2) the design of electoral systems and their influences on policy and party incumbency (Riker, William, Democracy in the United States, 2nd ed. [New York: Macmillan, 1965]Google Scholar for the U.S. Constitution; Brady, David and Mo, jongren, “Strategy and Choice in the 1988 National Assembly Election of Korea” [unpublished manuscript, Stanford University, 1991]Google Scholar for Korea; Bawn, Kathleen, “Electoral Incentives and Institutional Choice: The Development of the German Wahlgesetz, 1949–1953” [unpublished manuscript, Stanford University, 1991]Google Scholar for Germany); (3) the disenfranchisement of Southern blacks in the late nineteenth century (Woodward, C. Yann, Reunion and Reaction [Boston: Little, Brown, 1966]Google Scholar; Kousser, J. Morgan, Shaping of Southern Politics [New Haven, CT: 1974]Google Scholar; (4) the manipulation of registry laws by the Republicans to ensure their success over the Know-Nothings in the mid-1850s (Brady, David, Critical Elections and Congressional Policy Making [Stanford, CA: Stanford University Press, 1988]Google Scholar; Silbey, Joel, The Partisan Imperative [New York: Cambridge University Press, 1985Google Scholar; (5) more generally, efforts by Republican Progressives at the turn of the century to institute such reforms as the Australian ballot in an effort to destroy the city machines' ability to mass electoral support (Rusk, Jerrold G., “The Effect of the Australian Ballot Reform on Split–Ticket Voting: 1876–1908,” American Political Science Review 64 (1970): 1220–1238Google Scholar; Burnham, “Party Systems and the Political Process”); (6) the traditional literature on realignment (Chambers, William N. and Burnham, Walter Dean, eds., The American Parly Systems [New York: Oxford University Press, 1967]Google Scholar; McCormick, Richard, From Realignment to Reform [Ithaca, NY: Cornell University Press, 1981]Google Scholar; Clubb, Jerome M., Flanigan, William H., and Zingale, Nancy H., Partisan Realignment [Beverly Hills: Sage, 1980]Google Scholar; and (7) theoretical studies showing that even small differences in the aggregation rule lead to changes in electoral results (Cox, Gary, “Centripetal and Centrifugal Incentives in Electoral Systems,” American Journal of Political Science 34 (1990): 903–935)Google Scholar.

12. Nichols, Roy F., Blueprints for Leviathan (New York: Atheneum, 1963)Google Scholar.

13. Potter, David, The Impending Crisis (New York: Harper, 1976)Google Scholar; Weingast, Barry R., “Political Economy of Slavery,” (unpublished manuscript, Stanford University, 1991)Google Scholar.

14. Senate Republicans favored the admission of West Virginia by 21–7 while Democrats opposed it 1–5; the Unionists also opposed it by a 1–5 margin. Support or opposition within the Republican party on the vote did not appear to be associated with Radical sentiments, although some radicals (most notably, Charles Sumner) voted nay because West Virginia's constitution failed to abolish slavery immediately. Although West Virginia eventually became the one “new state” that could be counted on to send Democrats to Congress, its earliest representatives in Washington are labeled by Martis (p. 117) as “Unconditional Unionists” whose stance was “close to that of the range of sentiments within the Republican party and the Lincoln administration.” (p. 38) West Virginia's two Unconditional Unionist senators, Peter G. Van Winkle and Waitman T. Willey, later became Republicans when they acquired standard political labels (pp. 118–120). Martis, Kenneth C., Historical Atlas of Political Parties in the United States Congress, 1789–1989 (New York: Macmillan, 1989)Google Scholar.

15. Bensel, Richard, Yankee Leviathan (New York: Cambridge University Press, 1990), p. 88Google Scholar.

16. On strategies adopted in the House to overcome this Republican “executive cartel,” see Stewart, Charles III, “Lessons from the Post–Civil War Era,” in Cox, Gary and Kernell, Sam, eds., The Politics of Divided Government (Boulder, CO: Westview, 1991)Google Scholar.

17. For a discussion of related issues during the period immediately following that covered in this paper, see Bensel, Yankee Leviathan, pp. 88–101.

18. We also focus on statehood politics, understanding full well that there are other questions about the American electoral universe that we don't explore. The most obvious are black voting rights and matters such as the Force Bill. (See Bensel, Sectionalism.) We so proceed both because others have tackled black voting rights (Valelly, Richard, “Party, Coercion, and Inclusion: The Two Reconstructions and the South's Electoral Politics,” Politics and Society [forthcoming 1993])Google Scholar and because the statehood debate was a persistent matter whose affects on one important institutions, the Senate, can be traced fairly precisely.

19. Paullin, Charles Oscar, Atlas of the Historical Geography of the United States (Washington, D.C.: Carnegie Institution of Washington, 1932), plates 77 and 78Google Scholar.

20. “[Nevada's] population is obviously unworthy of the privilege of sending two men to the Senate, and has in fact allowed itself to sink, for all practical purpose, into sort of a rotten borough, which can be controlled or purchased by leaders of the Silver Ring.” (Bryce, James, American Commonwealth [New York, Macmillan, 1881]. Vol. II, p. 219)Google Scholar

21. Mormon historiography refers to this as the “Utah War.”

22. See Bensel, Yankee Leviathan, pp. 89–91.

23. The original Nevada Territory was roughly 60 percent its current size. It was extended to its current eastern borders in 1866, following the discovery of silver in the westernmost reaches of the Utah Territory, now the easternmost reaches of Nevada. At the same time the Nevada Territory was organized, the Colorado Territory was organized from land taken from the easternmost section of Utah and from Nebraska. On the creation of Nevada and its push to statehood, see Elliott, Russet R., History of Nevada (Lincoln: University of Nebraska Press, 1973)Google Scholar; Wren, Thomas, A History of the State of Nevada (New York: Lewis, 1904)Google Scholar; and Laxalt, Robert, Nevada: A Bicentennial History (New York: Norton, 1977)Google Scholar.

24. Sprague, Marshall, Colorado: A Bicentennial History (New York: Norton, 1976)Google Scholar.

25. Statements to the effect that the only reason Nevada was admitted was to provide one of these benefits to Republicans appear in Wilson, Woodrow, Congressional Government (Boston: Houghton Mifflin, 1956 [1885])Google ScholarPubMed; Bryce, American Commonwealth; and Elliott, History of Nevada.

26. Nevada had a population of 6,857 in the 1860 census and 42,491 in the 1870 census. No reliable census was taken between the decennial years. Based on a linear interpolation between 1860 and 1870, Nevada's population would have been 21,111 in 1864; a polynomial interpolation of the census returns for I860, 1870, 1880, and 1890 would have the population at 26,254 in 1864.

27. An additional motivation for the early admission of Nevada may also been at work: the potential threat of secession by California and Nevada during the Civil War may have led to the latter's earlv admission as a concession to both.

28. “Ratio of representation” is the number of inhabitants needed for each seat in the House of Representation using the “Vinton method” of apportionment. See note 29.

29. At this time seats in the House of Representatives were apportioned using a method called the “Vinton method.” Under this method, the total population of the country was divided by the size of the House as determined in the reapportionment bill. The resulting “ratio of representation” was essentially the number of inhabitants needed for each seat in the House of Representatives. The ratio of representation was then divided into the population of each state to determine how many representatives they would be entitled to; fractions were discarded at this step. Then, any remaining representatives needing to be allocated were assigned to the states having the largest remainders from the prior step (Congressional Quarterly, Guide to Congress, 2nd ed. [Washington: Congressional Quarterly Press, 1983], p. 565). Under the reapportionment that followed the 1860 census, the “ratio of representation” was 127,381.

30. Congressional Globe, 3 March 1863, 37th Cong., 3d sess., p. 1510.

31. Ibid. On the question of Nevada's true population, Wade argued:

I am credibly informed by the Commissioner of the General Land Office, who has taken considerable pains to ascertain the number of inhabitants there, that it cannot be less at this time than forty-five thousand, and increasing with unexampled rapidity. The vast riches of the mining region there have invited a population that is fast spreading over the whole of that region in a ratio that is almost unexampled, and I venture to predict here that in one year from this time there will be more than one hundred and fifty thousand inhabitants in this Territory. (Congressional Globe 3 March 1863, 37th Cong., 3d sess., pp. 1510–1511)

Nevada's population did not reach 150,000 until 1950.

32. Once admitted, there was no guarantee that these states would remain in the Republican fold forever. In the long term, all did not necessarily go according to the plans of Wade and the other Radicals more generally. In the late nineteenth century, when the Republican coalition fractured along sectional lines, the currency issue, and to some extent, an urbanrural cleavage, Nevada departed regularly from the Republican fold. Thus, like many of the states admitted from 1860 to 1896, Nevada was admitted for short- and medium-term partisan gain.

33. The act admitting Nebraska passed over the veto of President Andrew Johnson. The enabling act for Colorado was passed on 3 March 1875, the last day of the Forty-third Congress—the day before the Democrats gained control of the House for the first time in fourteen years. There were two interesting, and perhaps unique, features of this enabling act. First, once the state constitution had been ratified, all that was needed was for President Grant to declare Colorado admitted. Standard practice had previously required state constitutions to be approved by Congress before admission became official. Thus, the Democratic Forty fourth House would not be given an opportunity to obstruct Colorado's admission after its constitution had been ratified by its voters. Second, the enabling act timed the state ratification referendum just in time for Colorado to participate in the 1876 presidential election. Most interestingly, the enabling act allowed the new Colorado legislature to select the state's three presidential electors for that race. A good source on enabling acts and new state constitutions is Federal and State Constitutions, Colonial Charters, and Other Organic Laws of the States, Territories, and Colonies. Francis Newton Thorpe, comp. (Washington, D.C., Government Printing Office, 1909).

34. Paxson, Frederic L., Recent History of the United States, rev. and enlarged ed. (Boston: Houghton Mifflin, 1929), pp. 198–199Google Scholar.

35. Congressional Record, 9 April 1888, p. 2836.

36. Ibid., p. 3037.

37. Ibid., p. 3003.

38. Ibid., p. 948.

39. Roberts, Calvin and Robens, Susan A., New Mexico (Albuquerque: University of New Mexico Press, 1988), p. 154Google ScholarPubMed.

40. Congressional Record 14 February 1889, p. 1906.

41. Why Cleveland signed the bill, rather than veto it or let it become law without his signature, is unclear. At this point we simply note that with Harrison about to take office, vetoing the bill would have gained Cleveland practically nothing. A standard explanation for Cleveland's signing the bill—along with an explanation of why some Democrats defected in the House in the first place to support it—was that he wished to boost Democratic prospects in these four new states by having a Democratic president sign their enabling acts, a weak, though common, explanation.

42. Utah presents an interesting case which we intend to explore in greater detail in future research. Although Utah was regularly one of the largest territories, its entreaties for statehood went unheeded for forty years. An obvious problem was intolerance toward the Mormons who occupied the territory, especially disapproval of the polygamy that was practiced in its early days. Still, consistent with the partisan theme we wish to emphasize, Republicans were much more troubled by polygamy in Utah than Democrats, and were much more likely to want to delay Utah's admission as a state. Still, Democrats were not equally willing to overlook the religion issue, and thus Utah was the last territory they were able to get admitted.

43. This description of the geographic distribution of partisanship is taken from data generously provided by Walter Dean Burnham and from Martis, Historical Atlas.

44. These figures are for the elections that occurred prior to the secession of the southern states.

45. Burnham, “Party Systems,” p. 297.

46. Estimates of the Republican plurality in 1880 range from 2,000 votes (Congressional Quarterly, Guide to U.S. Elections, p. 340) to 39,000 (U.S. Census Bureau, Historical Statistics of the United States [Washington, D.C.: Government Printing Office, 1975], p. 1073)Google ScholarPubMed, out of 8.9 million votes cast.

47. Democrats and Republicans were of equal strength in the Senate following the 1880 elections, but the Forty-seventh Senate was organized by the Republicans.

48. An extended discussion of the Hayes-Tilden Affair is beyond the scope of this paper. On the affair and the resulting Compromise of 1876, see Stanwood, Edward, History of Presidential Elections (Boston: Houghton Mifflin, 1898)Google Scholar and Woodward, Reunion and Reaction. The precarious partisan balance entering the 1876 election clearly prompted both parties to fraud and more legitimate legal maneuvers in order to pick up just a handful of marginal electoral votes. An unanswered question is whether the absence of the new Republican states would have reduced the ex ante expected margin of the vote sufficiently to have reduced the probability the parties would have tried to manipulate the final result so blatantly.

49. It is often forgotten that the plank on the first Republican platform which declared against “popular sovereignty” in the territories on slavery (1856) read as follows:

Resolved: That the Constitution confers upon Congress sovereign powers over the Territories of the United States for their government; and that in the exercise of this power, it is both the right and the imperative duty of Congress to prohibit in the Territories those twin relics of barbarism—Polygamy, and Slavery.

50. One interesting aspect of the religious dimension is that the Democratic party was not uniformly tolerant of Mormons and Catholics. Hence, Republicans could often thwart Democratic efforts to admit Utah and New Mexico by bidding away a few Democrats with nativist leanings.

51. Occasionally, a member of Congress would argue that there was no lower bound on the population of a new state—that any white American citizen had the right to live in an organized state. We ignore this argument because even in the most extreme case, Nevada, the vast majority of its supporters, and especially the committees on territories, argued vehemently that Nevada was admissible because its population was significantly greater than zero.

52. For instance, Stephen A. Douglas answered one of Lincoln's famous questions at Freeport by noting the following: “I hold it to be a sound rule of universal application to require a territory to contain the requisite population for a member of Congress, before it is admitted as a State into the Union.” (quoted in Basler, Roy P., The Collected Works of Abraham Lincoln [new Bunswick, NJ: Rutgers University Press, 1953], Vol. III, p. 50)Google Scholar

53. The estimated population is created by linearly interpolating from the adjacent decennial censuses; average House district is calculated by dividing the national population in the previous census by the total number of House members under the previous decennial reapportionment. This last figure is not exactly the “ratio of representation,” but it is very close.

54. Taking the boundaries as given meant the following, in practice: First, we did not question any territorial boundary alterations that occurred before a territory was admitted to statehood. Second, we retained the division of the Dakota Territory when the Dakotas entered the Union. Third, we retained the partition of Virginia to create West Virginia. This is a conservative strategy from our perspective, since enquiring into boundary changes would yield even fewer Republican states in the long run.

55. The only serious difficulty we had in building a hypothetical composition of Congress and the Electoral College was in figuring out how to assign partisanship to states that were admitted after they acquired the requisite population (South Dakota, Washington, Utah, and Oklahoma). We adopted the following rules for imputing partisanship onto these states. First, we assigned the partisanship of the territorial delegate to the hypothetical member of the House, since the two offices were functionally similar. Second, we assumed that the party winning the territorial delegate race in a presidential election year would have carried that state in the presidential race, since the territorial delegate race was the only territory-wide office open for election in all territories. Third, we used the composition of the territorial legislature to estimate the partisanship of the senatorial delegation of the states, in order to simulate senatorial election by a state legislature. If both chambers were controlled by the same party, we assumed that both senators would have been from the legislative majority parly. If both chambers were controlled by different parties, we then assigned one senator to each party. In practice we had no difficulty in implementing these assignments, except in the case of Oklahoma, which had acquired a considerable population before being organized as a territory or being allowed a territorial delegate. In the case of Oklahoma, we assumed that it was a thoroughly Republican stale for the entire period, except in the 1896 election, when we assigned the state's electoral voles to Bryan. While open to dispute, this is a conservative treatment of Oklahoma from the perspective of our counterfactual argument: Had it voted Democratic al least some of the time, it would have increased the Democratic bias of our counterfactuals relative to the actual pattern.

56. See Polsby, Nelson, “The Institutionalization of the U.S. House of Representatives,” American Political Science Review 62 (1968): 144–168CrossRefGoogle Scholar; Bensel, Sectionalism.

57. Clubb, Jerome M. and Traugott, Santa A., “Partisan Cleavage and Cohesion in the House of Representatives, 1861–1974,” Journal of Interdisciplinary History 7 (1977): 375–401Google Scholar; Poole, Keith T. and Rosenthal, Howard, “The Enduring 19th Century Battle for Economic Regulation: The Interstate Commerce Act Revisited” (unpublished manuscript, Carnegie Mellon University, 1991)Google Scholar.

58. See Wolf, Thomas M., “Congressional Sea Change: Conflict and Organizational Accommodation in the House of Representatives, 1878–1921” (Ph.D. dissertation, M.I.T., 1981)Google Scholar, chap. 2; Poole and Rosenthal, “The Enduring 19th Century Battle,” p. 8.

59. Downs, Anthony, An Economic Theory of Democracy (New York: Harper, 1957)Google Scholar.

60. Cox, “Centripetal and Centrifugal Incentives.”

61. This summary of tariff history is drawn from Perry, O. H., “Proposed Tariff Legislation since 1883,” Quarterly Journal of Economics 2 (1887): 69–79Google Scholar; Taussig, F. W., Tariff History of the United States, 8th ed. (New York: Putnam, 1931)Google Scholar; Goldstein, Judith, Ideas, Interests, and American Trade Policy (Ithaca, NY: Cornell University Press, 1991)Google Scholar; and O'Halloran, Sharyn, Politics, Process, and American Trade Policy (Ann Arbor: University of Michigan Press, forthcoming)Google Scholar.

62. A new party is defined as gaining control of the government when the following three conditions are met: when (1) a single party controls both houses of Congress and the presidency following a congressional (or congressional/presidential) election, (2) this party did not control the entire federal government following the prior congressional election, and (3) the last time a single party controlled the entire federal government, it was a different party than the one currently controlling it.

63. In 1842 tariff rates had been undergoing a decade-long decline under the Compromise Tariff of 1833.

64. This analysis is not to suggest that partisan control is the only factor influencing tariffs—far from it. For instance, Republicans in the 1870s and 1880s were willing to finetune tariff rates in the face of embarrassing surpluses which threatened to undermine support for the protective tariff among their marginal supporters. Nonetheless, party turnover as denned in footnote 2 was a necessary condition for a major tariff change during this period.

65. Studenski, Paul and Krooss, Herman E., Financial History of the United States (New York: McGraw-Hill, 1952)Google Scholar; West, Robert C., Banking Reform and the Federal Reserve, 1863–1923 (Ithaca, NY: Cornell University Press, 1977)Google Scholar; Timberlake, Richard H., The Origins of Central Banking in the United States (Cambridge: Harvard University Press, 1978)Google Scholar; Walton, Gary M. and Rockoff, Hugh, History of the American Economy, 6th ed. (San Diego: Harcourt, 1991)Google Scholar.

66. Friedman, Milton and Schwartz, Anna J., A Monetary History of the United States (Princeton, NJ: Princeton University Press, 1963)Google Scholar.

67. What follows is based on Gilligan, Thomas W., Marshall, William J., and Weingast, Barry R., “Regulation and the Theory of Legislative Choice: The Interstate Commerce Act of 1887,” Journal of Law and Economics 32 (1989): 35–62Google Scholar and , Gilligan, , Marshall, and , Weingast, “The Economic Incidence of the Interstate Commerce Act of 1887: A Theoretical and Empirical Analysis of the Short-haul Pricing Constraint,” Rand Journal of Economics 21: 189–210Google Scholar. Poole and Rosenthal (“The Enduring 19th Century Battle”) provide a different account, arguing that the underlying issues were more structured unidimensionally than the account which follows. Either description of the underlying policy preferences guiding voting on the Interstate Commerce Act is consistent with the institutional argument we develop here, however.

68. While some have read this as a bill to protect shippers and others as a bill to bolster the railroads' fragile cartels, these conclusions result from focusing solely on one of the two problems generated by railroads and ignoring the other. Both perspectives are needed to understand the 1887 Act and neither view alone provides an accurate picture. See Gilligan, Marshall, and Weingast, “Regulation and the Theory of Legislative Choice.”

69. “Regulation and the Theory of Legislative Choice”; “The Economic Incidence of the Interstate Commerce Act of 1887.”

70. As Fiorina (1986) notes, this is exactly the mechanism used three years later in the Sherman Antitrust Act.

71. Dewey, R., The Long and Short Haul Principle of Rate Regulation (Columbus: Ohio State University Press, 1935)Google Scholar; Gilligan, Marshall, and Weingast, “The Economic Incidence of the Interstate Commerce Act of 1887.”

72. While the post–New Deal experience with the Court prompts us to associate partisan labels with judicial ideology cautiously, prior to the New Deal, commentators did not have a problem in associating partisan labels with judicial philosophy. See, for instance,Wilson, Woodrow, Congressional Government (Boston: Houghton Mifflin, 1956 [1885]), p. 45Google ScholarPubMed.

73. This is based on the family of models developed in the following: Calvert, Randall, McCubbins, Mathew D., and Weingast, Barry R., “A Theory of Political Control and Agency Discretion,” American Journal of Political Science 33 (1990): 588–611Google Scholar; Peter H. Lemieux and Charles Stewart III, “Advice? Yes. Consent? Maybe: Senate Confirmation of Surpreme Court Nominations” (presented at the annual meeting of the American Political Science Association, Washington, D.C.); , Lemieux and , Stewart, “Senate Confirmation of Supreme Court Nominations from Washington to Reagan,” Hoover Institution Working Paper Series, P-90–3, 1990Google Scholar.

74. On the courts, see Lemieux and Stewart, “Advice? Yes. Consent? Maybe,” and “Senate Confirmation of Supreme Court Nominations.” On the National Labor Relations Board (NLRB), see Moe, Terry M., “Interests, Institutions, and Positive Theory: The Politics of the NLRB,” Studies in American Political Development 2 (1987): 236–299CrossRefGoogle Scholar.

75. The scoring was developed as follows: Justices appointed by Democrats and confirmed by Democratic Senates were given a score of 1, justices appointed by Democrats and confirmed by Republican (or Whig) Senates were given a score of 0.5, justices appointed by Republicans and confirmed by Democratic Senates were given a score of –0.5, and justices appoited and confirmed by Republicans were given a score of –1. The score for the entire Court at any given time is simply the sum of the justices' scores. President Tyler was considered a Democrat for this exercise, although this does not change the figure significantly. The short-term blips in Figure 5 are generally periods in which a vacancy occurred on the Court.

76. In order to calculate this hypothetical partisan composition of the Court, we took the pattern of Court vacancies as given by the historical record. This is problematic for at least two reasons. Most obviously, a different set of judges appointed would have given rise to a different pattern of retirements due to the idiosyncracies of health, career ambitions, and the like. More interestingly, the retirement pattern as it exists is not random. See King, Gary, “Presidential Appointments to the Supreme Court: Adding Systematic Explanation to Probabilistic Description,” American Politics Quarterly 15 (1987): 355–372Google Scholar. We would not be surprised to discover that justices retired strategically, perhaps to influence the subsequent composition of the Court. For instance, Republican justices may have been more willing to retire when they knew a Republican would get to replace them and a Republican Senate would handle the confirmation. An alternative history with more Democratic Senates in the 1880s and 1890s may have slowed down the retirement rate of the Republican Justices who were appointed in the 1860s and 1870s. On the politics of Supreme Court appointments during this period, see Abraham, Henry J., Justices and Presidents, 2nd ed. (New York: Oxford University Press, 1985)Google Scholar.

77. Lochner v. Sew York, 198 U.S. 45.

78. See Keller, Morton, Affairs of State (Cambridge: Harvard University Press, 1977)CrossRefGoogle Scholar.

79. While positive political theory remains a relative newcomer to the study of macropolilical events, it may well prove the most valuable tool for explaining the institutional foundations of such events. See Riker, William H., Liberalism Against Populism (San Francisco: Freeman, 1982)Google Scholar, Poole, Keith T. and Rosenthal, Howard, “Patterns of Congressional Voting,” American Journal of Political Science 35 (1991): 228–278Google Scholar, and Weingast, Barry R., “The Political Economy of Slavery” (unpublished manuscript, Stanford University, 1991)Google Scholar.