INTRODUCTION

A Census Office was not formally established until 1850, after the duty of taking the census had been transferred from the Department of State to the Department of the Interior. The Office, however, was temporary; it was established for each decennial census and disbanded when work on the census was completed. Effective July 1, 1902 (5 Stat. 51), the Census Office became a permanent office, but the terms Census Bureau and Census Office were used interchangeably for some time thereafter. The name Bureau of the Census was applied to the permanent office by the Secretary of Commerce and Labor in 1903 but was not formally fixed by legislation until 1954. In this inventory, however, the term Census Office is used for the temporary clerical staffs or offices set up every 10 years from 1790 to 1900, and the term Bureau of the Census or Census Bureau is used for the permanent office established in 1902.*

The necessity for census taking had its origin in a compromise of the Constitutional Convention, which agreed that each State should have two Senators in Congress and that Representatives should be apportioned among the States according to population. Section 2 of Article I of the Constitution consequently provided that Representatives and direct taxes be apportioned among the States according to a population count to be conducted within 3 years after the first meeting of Congress and every 10 years thereafter "in such Manner as they shall by Law direct.'

An act of March 1, 1790 (1 Stat. 101), provided for taking the first census in the States that had ratified the Constitution at that time. The provisions of the act were extended to include Rhode Island on July 5, 1790 (1 Stat. 129), and the newly admitted State of Vermont on March 2, 1791 (1 Stat. 197). The original act of March 1, 1790, with minor modifications and extensions, governed the taking of the census until 1850. It went beyond the constitutional requirement, which excluded Indians not taxed; and it distinguished free persons, including those bound to service for a term of years, from all others. Its schedule of inquiries

See the table at the end of this introduction, "Officials Who Directed the Taking of the Census From 1790 to Date.' In congressional reports the terms Census Bureau and Census Office were used interchangeably to refer to the proposed permanent census organization. In legislation the terms Census Office and Bureau of the Census were used interchangeably to refer to the permanent census organization until section 3 of an act of August 31, 1954 (68 Stat. 1025), provided that references in other laws to the Census Office should be considered applicable to the Bureau of the Census. name of the Census Office was changed to the Bureau of the Census on July 1, 1903, by an order of the Secretary of Commerce and Labor, but the legality of the new name was questioned by the Director of the Census for several years thereafter (see Record Group 40, General Records of the Department of Commerce, file 20873).

also included the names of the heads of families, distinguished the number of free white males 16 or more years old from those under that age, and listed the numbers of free white females and of other persons. This act gave responsibility for the enumeration to the marshals of the U.S. judicial districts, but it empowered no central authority to supervise them in carrying out their duties. A provision for taking the second census

by the marshals "under the direction of the Secretary of State" was included in an act approved February 28, 1800 (2 Stat. 11), but only after considerable debate in the House.

The taking of the decennial enumerations remained under the jurisdiction of the Secretary of State through the sixth census (1840), when the first phase of census taking came to a close. The characteristics of this phase may be summarized as follows: the family rather than the individual was the unit of data; there was no tabulation beyond the simple addition of the entries sent in by the marshals; there was no attempt to present details uniformly by cities and towns or to summarize returns for each State by county unless the marshal had done so; and there were numerous errors in compiling and summarizing the data.

Long before the time of the seventh census Members of Congress, as well as statisticians and other scholars both within and outside the Federal Government, had wanted to go beyond the constitutional requirement for an enumeration and to use the decennial censuses as a means of collecting extensive statistical data. In the House debates on census bills, as reported in the Annals of Congress, motions to collect occupational statistics in the first census and to gather statistics on the "number of dwelling houses" in the fourth census had been defeated. Memorials of the Connecticut Academy of Arts and Sciences and of the American Philosophical Society, presented to the Senate on January 10, 1800 (Annals of Congress, 6th Cong., 1st sess., p. 25), had suggested the collection of data on agriculture, occupations, naturalization, and citizenship. As a result of this urging to expand the decennial enumeration the legislation authorizing the third and fourth censuses (acts of May 1, 1810, and March 14, 1820) included provisions for taking a census of manufacturing establishments (2 Stat. 605, sec. 2; 3 Stat. 553, sec. 10). Later, on April 6, 1836, Francis Lieber, professor of history and political economy at South Carolina College, presented a memorial to Congress recommending that statistics of "a comprehensive nature" be compiled by "a gentleman of established literary reputation, and acknowledged ability" (24th Cong., 1st sess., S. Doc. 314). The sixth census act, March 3, 1839 (5 Stat. 336), provided for the collection of statistics pertaining to "the pursuits, industry, education, and resources of the country.

The texts of these memorials appear in 41st Cong., 2d sess., H. Rept. 3, Jan. 18, 1870, p. 35-36. In American State Papers: Miscellaneous, 1:202-203, the text of the American Philosophical Society's memorial is dated January 23, 1800.

As the time for the seventh census approached, Jesse Chickering and Nahum Capen, members of the 10-year-old American Statistical Association, suggested that census taking be put in the hands of "persons of known

ability in the science of statistics" and recommended the appointment of a board of commissioners "without reference to party" to study the matter (30th Cong., 2d sess., S. Misc. Doc. 64). As a result a Census Board composed of the Secretary of State, the Attorney General, and the Postmaster General was established by section 2 of an act of March 3, 1849 (9 Stat. 403), to prepare schedules and forms for the seventh census. Joseph C. G. Kennedy, who was clerk or secretary of the Board from May 1, 1849, to May 31, 1850, and who later became Superintendent of the Census, consulted with statisticians, notably Lemuel Shattuck, whose proposal for recording the name and description of every person enumerated had been adopted in the Boston census of 1845.

Shattuck's proposal, which was incorporated in the seventh census schedule for free inhabitants, ushered in the second phase of census taking and introduced the present method of recording facts concerning each individual separately. The scope of the data collected was further extended by the use of six schedules: of free inhabitants, slave inhabitants, mortality, products of agriculture, products of industry, and "social statistics.' These schedules, which were prepared and printed by the Census Board, were incorporated in an act of May 23, 1850 (9 Stat. 428), providing for the taking of the seventh and subsequent censuses. Because the census function had been transferred from the Secretary of State to the Secretary of the Interior when the Department of the Interior was established (9 Stat. 395), section 19 of the act of May 23, 1850 (9 Stat. 431), required the Secretary of the Interior to "carry into effect the provisions of this act." To discharge this duty the Secretary was to appoint "a suitable and competent person" as superintending clerk and such subordinate clerks and officers as were necessary. The superintending clerk, later called the Superintendent of the Census, was to have "the general management of matters appertaining" to the census.

By 1870 the second phase of census taking was drawing to a close. In January of that year Representative James A. Garfield of the House Select Committee on the Ninth Census submitted a report (41st Cong., 2d sess., H. Rept. 3, January 18, 1870), pointing out the defects of the 1850 census act. That law was no longer adequate because the basis of popular representation had been changed by the 13th and 14th amendments to the Constitution, the gathering of statistics had become more scientific, and data should be collected to show the "astonishing development of . . . the resources of our country." The report included a draft of a proposed bill for taking the ninth census, which passed the House but was defeated in the Senate. This bill, however, laid the groundwork for the 10th census act. The accompanying report recommended that the U.S. marshals should no longer take the census because they were not chosen "with a view to their fitness for census taking, or any statistical inquiry" and because the congressional districts of the States

were more suitable as enumeration districts than were the judicial districts. An act of March 3, 1879 (20 Stat. 473), therefore, providing for the taking of the 10th and subsequent censuses, authorized the President to appoint supervisors of the census and provided that the supervisors should recommend appropriate subdivisions of their districts and suitable persons to become enumerators. The enumerator's oath for the first time forbade unauthorized disclosure of census information. This act also authorized the Census Office to withdraw certain kinds of statistical inquiries from the enumerators and put them in the hands of experts and special agents. This act was the model for the act of March 1, 1889 (25 Stat. 760), which provided for taking the 11th census.

The 1880 and 1890 censuses constitute the third phase of census taking. The number and variety of census subjects and the change in methods of collecting the statistics resulted in more complete and accurate returns for these censuses than had been available for their predecessors. On the other hand the simultaneous collection of so many returns, as well as the variety and complexity of the schedules, prolonged the decennial census period for the 11th census until 1897.

The fourth phase of census taking, which resulted finally in a permanent census bureau, stemmed from early suggestions that census duties be assigned to a proposed bureau of statistics. Such suggestions, made on June 17, 1844, by the Select Committee on Statistics of the House of Representatives (28th Cong., 1st sess., H. Rept. 579) and by the Secretary of the Interior in his annual reports, 1861-65, came to naught, however, when the Bureau of Statistics was established not in the Interior but in the Treasury Department on July 28, 1866 (14 Stat. 330). In the meantime the Superintendent of the Seventh Census had pointed out the disadvantages of disbanding the Census Office staff at the end of each decennial census period as soon as it "acquires familiarity with statistics, and is educated to accuracy" (Compendium of the Seventh Census, 1854, p. 18-20). A solution to this problem was suggested by Gen. Francis A. Walker, Superintendent of the Census for the 9th and 10th censuses, in an article in the January 1888 issue of the Quarterly Journal of Economics. Walker recommended limiting the decennial census to the fields of population and agriculture and transferring the census function to the existing Bureau of Labor in the Department of the Interior. Carroll D. Wright, Commissioner of Labor, expressed substantial agreement with Walker in an article in the November 1891 issue of Popular Science Monthly that recommended the establishment of a permanent census office.

*The Bureau of Labor was established in the Department of the Interior on June 27, 1884 (23 Stat. 60). The Bureau was merged with a newly established Department of Labor (also in the Interior Department) on June 13, 1888 (see secs. 1 and 9, 23 Stat. 182, 183). The head of the Bureau and the Department was known as the Commissioner of Labor.