Government

Y 1-26. Methods of electing presidential electors, 1788-1836.

Source: Charles O. Paullin, Atlas of the Historical Geography of the United States, Carnegie Institution of Washington and American Geographical Society of New York, 1932, p. 89 (courtesy of the Carnegie Institution).

The presidential electors of each State, now chosen by popular vote in all States, are selected, according to the Constitution, “in such manner as the legislature thereof may direct." The development of political party direction of the electoral college was not anticipated in the Constitution and, during the early years of the Republic, electors were chosen in the several States by a number of different devices. The principal methods were election by the State legislature itself, by State electors popularly chosen to elect presidential electors, and by direct popular vote for the electors. With few exceptions, presidential electors have been elected by popular vote since 1828. The Legislature of South Carolina, however, continued to elect presidential electors until 1860. Since the Civil War, legislatures have chosen electors only twice-in Florida in 1868 and in Colorado in 1876.

Y 27-78. Voter participation in presidential elections, by State, 1824-1968.

Source: Walter Dean Burnham, Dept. of Political Science, Massachusetts Institute of Technology, unpublished data. The explanatory notes which follow were prepared by Professor Burnham.

The United States, unlike some other countries, has never developed an automatic, governmentally-operated system for enrolling potential voters. The uniform practice since the earliest times has been that each of the States is the sole judge of the electoral procedures which it prescribes within its jurisdiction, subject only to constitutional amendments, congressional legislation enacted pursuant to such amendments or other portions of the Constitution, and Federal judicial decisions. All of these may limit or abolish the States' discretion in specified areas of legal procedure pertaining to elections; otherwise, the general rule stated here has continuously applied to the conduct of elections in the United States.1

1 Only the most salient of such interventions are mentioned here, specifically as they apply to the composition of the potentially eligible electorate. They are: The Reconstruction Act of March 2, 1867, which (with the Supplemental Act of March 23, 1867) required that the ten ex-Confederate States still without Federal representation eliminate all racial barriers to the suffrage as a precondition for readmission.

The Fifteenth Amendment to the Constitution (1870), which forbade either the States or the Federal Government to deny the right of citizens to vote "on account of race, color, or previous condition of servitude," and subsequent implementing legislation of 1870, 1871, and 1875. (See also Guinn v. United States, 238 U.S. 347 (1915), in which the Supreme Court struck down the "grandfather clause" as a patent attempt to evade the command of this Amendment.)

The Nineteenth Amendment to the Constitution (1920), which enfranchised women on the same constitutional terms as men.

The Supreme Court's invalidation of the white primary, and of the doctrine that political parties, being private associations, could exclude Negroes or any others they chose to exclude from their nominating processes. (Smith v. Allwright, 321 U.S. 649, (1944).)

The Twenty-third Amendment to the Constitution (1961), extending presidential suffrage to the District of Columbia.

The Twenty-fourth Amendment to the Constitution (1964), prohibiting the levying of a poll tax or any other tax as a prerequisite to voting in Federal elections. (See also the Supreme Court's extension of this prohibition to State and local elections as well, Harper v. Virginia State Board of Elections, 383 U.S. 663 (1966).)

The Civil Rights Act of 1965 suspending all literacy tests in Alabama, Alaska, Georgia, Louisiana, Mississippi, South Carolina, Virginia, 40 counties in North Carolina, and one county in Arizona, and establishing Federal registrars in the affected areas. The Civil Rights Act of 1970, which continued suspension for another five years; extended it to all literacy tests; lowered the minimum voting age from 21 to 18 in all Federal and State elections; and lowered the minimum residence requirement for voting in presidential elections to a uniform 30 days.

The Twenty-sixth Amendment to the Constitution (1971) which formally reaffirmed the lowering of the voting age minimum from 21 to 18 years.

As a result of this heterogeneity, it is not possible to achieve precise statements of the eligible electorate. The data in series Y 27-78 are estimates, and should be read throughout with that point in mind.

Every estimate of voting participation is, in effect, a ratio between a numerator and a denominator. Errors in estimates may occur because of errors in the numerator, the denominator, or both. Problems with the numerator include the following: (1) The reported vote cast may have been heavily inflated by fraudulent ballot-box stuffing, as for example in Plaquemines Parish, Louisiana, in 1844, or in Kansas City, Missouri, between 1934 and 1938; (2) the stated vote may be only a fraction of the real vote cast, either because of fradulent suppression of returns, other forms of pressure on the electorate, or failure of subdivisions to report within the legal time limit (the latter very frequently occurring in Texas); (3) more or less major compiling or reporting errors, without fraudulent intent, may exist, and clearly did in a number of cases in the nineteenth century; and (4) available returns may be significantly fragmentary because the original records were lost.

Problems with the denominator, i.e. the population base, relate primarily to its derivation. Information which has been compiled into the denominator falls into the following classifications: (1) Age cohorts; (2) sex by age for and following every point at which women were enfranchised by State law or by the Nineteenth Amendment; (3) race, which substantially means adding the Negro adult male population to the denominator base in 1868 (ten Southern States) or 1870; (4) citizenship status, reported separately as "male citizens" in 1870, and in more detail beginning with the 1900 census.

The following possible elements of a denominator have not been compiled, though they have entered into State legislation regulating eligibility to vote: (1) Literacy of the adult male/adult population; (2) taxpaying components of the adult male/adult population; (3) other components (for example, the total number registered) which define those legally entitled to vote at any given time. There are two reasons for such exclusions: Many of such devices were deliberately employed, particularly between 1890 and shortly after 1960, to violate the letter and the spirit of the Fifteenth Amendment; and they are extremely heterogeneous-for example, some States have registration reporting covering all jurisdictions, some do not, and the times for which such information is available are extremely diverse. The four major components of the denominator, the estimated eligible population, alone can be developed more or less accurately for all States and time periods but even these have very significant problems. In general, the denominator estimates are much more precise from 1900 on than they are for years prior to that date. The reason for this lies in the changing nature of census reporting of critical components which enter the calculation.

Age cohorts. From 1870, the adult male adult population is specifically enumerated so that the whole number of those 21 years old and over can be stated. This was not the practice earlier. Procedures followed here were to sum all white male age cohorts entirely above 21, and to add to that sum a fraction of any age cohort which bracketed that age, i.e., which included both males above and below 21 years. The fraction was derived by a simple division of the relevant bracketed age grouping, and addition of the quotient to the sum already derived. For example: The 1820 census yields free white male age classifications of 16–25, 26–44, and 45 and over. For Maine in that year, the total number of free white males 26 and over was