¹ Ames, H. V., “Proposed Amendments to the Constitution,” H. Doc. 353, Pt. 2, 54th Cong., 2nd sess., pp. 77–123Google Scholar; Musmanno, M. A., “Proposed Amendments to the Constitution,” H. Doc. 551, 70th Cong., 2nd sess., pp. 44–51, 60–64Google Scholar; Tansill, Charles C., “Proposed Amendments to the Constitution, S. Doc. 98, 69th Cong., 1st sess., p. 143Google Scholar; Loeffler, Carl A., “Proposed Amendments to the Constitution,” Senate miscellaneous publication, p. 111Google Scholar [Y. 1.3:C76/4/926–47(1949)].

^1a S. J. Res. 2, favorably reported by the Senate Committee on the Judiciary, S. Rept. 602, 81st Cong., 1st sess. Cong. Rec. (Daily), Vol. 96, pp. 854, 908–920, 971, 981–985, 987, 1087–1095, 1098–1101, 1108–1110, 1176–1191, 1289–1307. Forty-six Democrats and eighteen Republicans voted for the resolution; and four Democrats and twenty-three Republicans voted against it.

² Identical or almost identical resolutions offered in the House are H. J. Res. 2 by Gossett (D., Tex.), H. J. Res. 10 by Cannon (D., Mo.), H. J. Res. 11 by Celler (D., N. Y.), H. J. Res. 51 by Davis (R., Wise), H. J. Res. 78 by Hays (D., Ark.), H. J. Res. 81 by Johnson (R., Cal.), H. J. Res. 82 by Priest (D., Tenn.), and H. J. Res. 121 by Boggs (D., La.). H. J. Res. 2 was reported favorably by the House Committee on the Judiciary, H. Rept. 1011, 81st Cong., 1st sess.

³ New York Times, Dec. 7, 14, and 16, 1948. An Alabama law (Act #386, approved July 7, 1945) instructing electors to cast their ballots for the nominees of the party's national convention was declared unconstitutional by the Alabama Supreme Court. Op. of the Justices, 34 So. 2d 598 (1948). The Alabama Supreme Court also refused injunctive relief to Adcock et al. who sought to compel the electors to cast their ballots for Truman. The Supreme Court of the United States refused to review the decision of the Alabama Supreme Court. Adcock et al. v. Albritton et al., 335 U.S. 887 (1948). The Supreme Court of the United States also denied motions for leave to file petitions for injunction and mandamus. Adcock et al. v. Albritton et al., 335 U.S. 882 (1948); Folsom et al. v. Albritton et al., 335 U.S. 882 (1948).

⁴ Wilmerding, Lucius Jr., “Reform of the Electoral System,” Political Science Quarterly, Vol. 64, pp. 1, 17 (Mar., 1949)CrossRefGoogle Scholar.

⁵ Ibid., pp. 16–18, 20–21.

^5a The Senate adopted an amendment proposed by Senator Lucas. Among other things, the amendment would empower Congress to provide by law for the case of the death of any persons from whom Congress may choose a President or Vice-President whenever the right of choice shall devolve upon Congresss. Cong. Rec. (Daily), Vol. 96, p. 1304Google Scholar. Presumably this does not cover the case of a candidate who dies after receiving the requisite electoral vote. Such a case, it seems, would fall within the provisions of the Presidential Succession Act. 61 Stat. L. 380 (1947)Google Scholar.

⁶ These and all other calculations for the 1948 election, unless otherwise indicated, are the author's and are based on the statistics compiled by the Associated Press from the official canvassing boards in the forty-eight states. New York Times, Dec. 11, 1948.

^6a Cong. Rec. (Daily), Vol. 96, pp. 1180–1191, 1289–1307Google Scholar. The House minority report suggested requiring an electoral plurality of 35 percent. Op. cit., p. 27. Senator Ferguson suggested in his minority report that provision should be made for election by a joint session of Congress—voting as individuals and not as states—in case no candidate received the required number of electoral votes. S. Rept. 602, Pt. 2, 81st Cong., 1st sess., p. 2.

⁷ Kallenbach, Joseph E., “Presidential Election Reform,” Cong. Rec. (Daily), Vol. 95, pp. 4531–4536, n. 27 (Apr. 13, 1949)Google Scholar. Statistics of a similar kind can be found throughout Hearings before a Subcommittee on the Judiciary, United States Senate, 81st Cong., 1st sess., on S. J. Res. 2; and Hearings before Subcommittee No. 1 of the Committee on the Judiciary, House of Representatives, 81st Cong., 1st sess., on H. J. Res. 2 (serial 3).

⁸ Calculations based on United States Bureau of Census estimate of population, July 1, 1947. Of course, considerable shifts of population have taken place since the census of 1940. Information Please Almanac 1948 (New York, 1947), p. 266Google Scholar.

⁹ In 1948, for example, if the election had been determined by the nation-wide popular vote regardless of state lines, 17 states would have had their relative strength increased, and 31 states would have had their relative strength decreased. California would have had her voice in electing the President increased 75.47%, Illinois (55.21%), Massachusetts (43.68%), New York (43.37%), Indiana (38.97%), New Jersey (32.92%), Ohio (28.10%), Washington (23.36%), Michigan (21.13%), Minnesota (20.17%), Connecticut (19.24%), Pennsylvania (16.42%), Wisconsin (16.06%), Missouri (14.80%), Iowa (13.22%), Kansas (7.50%), and West Virginia (2.06%). The following states would have had their voice in choosing the President decreased: South Carolina (80.56%), Alabama (78.67%) Nevada (77.35%), Mississippi (76.70%), Arkansas (70.62%), Wyoming (63.19%), Georgia (61.95%), Virginia (58.45%), Vermont (55.22%), Louisiana (54.59%), Arizona (51.66%), Tennessee (50%), Delaware (49.38%), New Mexico (49%), Texas (45.58%), Maine (42.25%), Idaho (41.43%), North Dakota (39.84%), Montana (38.78%), North Carolina (38.38%), New Hampshire (36.92%), South Dakota (31.74%), Utah (24.57%), Oklahoma (2130%), Florida (21.23%), Maryland (18.65%), Kentucky (18.44%), Nebraska (11.15%), Rhode Island (11.02%), Colorado (6.37%), and Oregon (4.69%).

These figures indicate that a constitutional amendment making electoral strength proportional to popular vote could not secure ratification in the requisite 36 states.

^9a Jackson received only 37.9 percent of the electoral vote. Thus, under the Lucas amendment, the choice of a President would have devolved upon Congress.

¹⁰ Electors were appointed by the legislature in New York, Delaware, South Carolina, Georgia, Vermont, and Louisiana. Electors were chosen by popular vote in districts in Maryland, Kentucky, Tennessee, Illinois, Maine, and Missouri. In other words, the general ticket system was used in only twelve states. Historical Statistics of the United States (Washington, 1949), p. 288Google Scholar. In any case, Jackson only polled 43 per cent of the popular vote, and his plurality was only 44,804. Ibid., p. 290. At the time these computations were made, the Lodge plan provided that calculations were to be carried to three decimal places “unless a more detailed calculation would change the result of the election.” Consequently, these computations were carried to five decimal places. Later the Senate adopted an amendment limiting calculations to three decimal places. In this case, however, the two extra decimal places make no material difference.

^10a The votes in the House were 87 for Adams, 71 for Jackson, and 54 for Clay. It is nothing more than optimistic speculation to assume that the 54 votes for Clay and the 48 senatorial votes would have been cast so as to have given Jackson the requisite 131 votes.

¹¹ Haworth, Paul L., The Hayes-Tilden Disputed Presidential Election of 1876 (Cleveland, 1906)Google Scholar; U.S. Electoral Commission, The Electoral Count of 1877 (Washington, 1877)Google Scholar. Not only was there no popular vote in Colorado because the electors were appointed by the legislature, but also there was so much fraud that nobody can determine exactly what percentage of the popular vote each candidate had. No official count of the popular vote was made. According to the Democratic count, Tilden had a plurality of 264,292; but according to the Republican count, his plurality was only 252,224. Application of the Lodge formula to the Democratic count would give him a margin of 11.2 electoral votes; and under the Republican count he would have had a margin of 10.5 electoral votes. For the two counts, see Stanwood, Edward, A History of the Presidency (Boston, 1904), Vol. 1, p. 383Google Scholar.

¹² Elections would become closer contests as a result of dividing the electoral vote in each state. For example, in the election of 1908, the electoral vote margin would have been 3.9; in the election of 1900, there probably would have been a margin of about 0.1 electoral votes. Hearings on H. J. Res. 2, op. cit., p. 99. Moreover, this would make a few hundred popular votes, properly distributed, the determinant of the outcome. See n. 16, below.

¹³ According to calculations prepared by the Legislative Reference Service of the Library of Congress, Cleveland would have garnered 202.9 electoral votes instead of 168, and Harrison would have received only 185.8 instead of 233. Hearings on H. J. Res. 2, op. cit., p. 99.

¹⁴ Republicans: popular vote, 4,449,653 or 48.3 percent; electoral vote, 214; electoral vote under Lodge formula, 175.1. Democrats: popular vote, 4,442,035 or 48.2 percent; electoral vote, 155; electoral vote under Lodge formula, 181.9. Ibid., p. 99.

¹⁵ Republicans: popular vote 7,035,638; electoral vote 271; electoral vote under Lodge formula, 215.3. Democrats: popular vote, 6,467,946; electoral vote 176; electoral vote under Lodge formula, 221.3. Ibid., p. 99.

¹⁶ According to the calculations of the Legislative Reference Service, McKinley would have had a margin of one-tenth of an electoral vote. Republicans: popular vote, 7,219,530 or 51.7 percent; electoral vote 292; electoral vote under Lodge formula, 217.3. Democrats: popular vote, 6,358,071 or 45.5 percent; electoral vote, 155; electoral vote under Lodge formula, 217.2. Ibid., p. 99.

According to the author's calculations based on statistics in the Statistical Abstract of the United States 1908, p. 42, which seem to be the same statistics used by the Legislative Reference Service, McKinley would have had a pluraltiy of 0.192 electoral votes.

According to Senator Taft's calculations, Bryan would have had a plurality of 4.3 electoral votes. The statistics on which Taft based his computations are as follows: Republicans' popular vote of 52.8 percent and a plurality of 861,000; Democrats' popular vote of 47.2 percent. Cong. Rec. (Daily), Vol. 96, pp. 1298, 1300–1301Google Scholar.

A shift of 100 votes from McKinley to Bryan in each of the eleven states of the solid South and in Kentucky would have given Bryan a margin of .002 electoral votes (author's calculations based on statistics in the Statistical Abstract of the United Slates 1908, p. 42). Yet McKinley would still have had a popular margin of 859,059 and an absolute majority of 51.69 percent. Election returns will actually vary by several thousand votes as recounts and more careful calculations progress. For example, according to the New York Times compilation for the 1948 election (op. cit.), Truman had a plurality of 2,135,336. But according to returns corrected to January 29, 1949, hia plurality had increased by 1,189 (H. Rept., op. cit., pp. 14–15). Certainly, these figures suggest that the Lodge formula would give rise to contested returns in a close election.

¹⁷ The author's calculations based on the statistics used by the Legislative Reference Service, Hearings on H. J. Res. 2, op. cit., p. 99. It will be noticed that the author used the electoral vote calculations of the Legislative Reference Service for the election of 1900.

¹⁸ For the Democrats, the rank-difference coefficient of correlation is positive 114; for the Republicans, the rank-difference coefficient of correlation is negative .063, which is so small that it could not possibly account for the Republican disadvantage.

¹⁹ The calculations in Table II are based on statistics in the following sources: Stanwood, op. cit., Vol. 1, pp. 417, 448; and the Statistical Abstracts of the United States, 1912 (pp. 727–735), 1917Google Scholar (pp. 698, 702), 1920 (pp. 758, 762), 1924 (pp. 141–142), 1928 (pp. 166–167), 1933 (pp. 155–156), 1937 (pp. 159–160), 1941 (pp. 174–175), 1944–45 (pp. 251, 253). The “solidness of the South” is an average percentage of the popular vote polled by the Democrats in the eleven states, weighting each state according to its electoral vote. This method was used because it makes a great difference, for example, whether the Democrats poll 90 percent of the popular vote in a state with three electoral votes or in a state with twenty electoral votes. Of course, the eleven states of the Solid South are: Alabama, Arkansas, Florida, Georgia, Louisiana, Mississippi, North Carolina, South Carolina, Tennessee, Texas, and Virginia.

The rank-difference coefficient of correlation between the solidness of the South and the Democratic advantage under the Lodge formula is .680.

²⁰ Popular participation in the South in relation to popular participation in the rest of the country is measured in terms of the relation between the percent of the national popular vote polled in the South and the percent of the total electoral vote cast by the South. In 1928, for example, the South cast 23.729 percent of the electoral vote (126 out of 531); but only 9.164 percent of the nation-wide popular vote was polled in the South. Of course, 9.164 is 38.619 percent of 23.729. Thus popular participation of that year was figured as 38.619 percent. In other words, if the election had been determined by the nation-wide popular vote, the South would have had only 38.619 percent of the voice which the eleven states actually had in electing the President. It will be noted that the low popular participation that year partially offset the lack of southern solidarity.

The rank-difference coefficient of correlation between the Democratic advantage under the Lodge formula and popular participation in the South is negative .587; the coefficient of multiple correlation between Democratic advantage and solidness of the South and popular participation in the South (R_1.23) is .701.

²¹ The Democratic disadvantage in 1892 cannot be explained entirely by the relatively high popular vote in the South and the relative lack of southern solidarity. Another important factor was the Populist invasion of Democratic strength in the small western states which are over-represented in the electoral college:

Calculations based on election returns in the Statistical Abstract of the United States 1912, pp. 728, 735.

²² See n. 17, supra. The author used the statistics of the Legislative Reference Service for the election of 1876.

^22a Calculations carried to five decimal places.

²³ The author's calculation is based on the statistics in Edgar Eugene Robinson, They Voted for Roosevelt (Stanford, 1947), pp. 42–46Google Scholar. The popular vote in this hypothetical case is as follows: Roosevelt, 41.7%; Hoover, 55.4%; others, 2.4%.

²⁴ Hearings on S. J. Res. 2, op. cit., p. 88.

²⁵ Ibid., pp. 8–9, 71–90; Cong. Rec. (Daily), Vol. 95, pp. 2946–2951 (Mar. 22, 1949)Google Scholar.

^25a For example:

Source: Statistical Abstract of the United States 1943, pp. 234, 237; ibid., 1944–45, pp. 251, 257.

²⁶ On this point, see the minority report on H. J. Res. 2, op. cit., pp. 28–29.

²⁷ Binkley, Wilfred E., American Political Parties (New York, 1943), p. 317Google Scholar.

²⁸ Hearings on S. J. Res. 2, op. cit., p. 209.