All these six censuses had a common defect which robbed them of ultimate statistical value: They lacked uniformity in method, and to a large extent could not furnish comparisons or accurate deductions. The enumeration ex tended over long periods, many months at a time; the results were not summarized by coun ties, nor uniformly by cities or towns; the clas sification of ages of colored people was on a different basis from that of whites, etc.
The seventh census, that of 1850, begins a new era. The blunders of the last one, and the public disgust and indignation, had aroused great discussion as to proper methods; and on 3 March 1849, the same act which created the Department of the Interior also created a cen sus board, consisting of the Secretary of State, the Attorney-General and the Postmaster-Gen eral, to prepare a census schedule including not above 100 inquiries. They framed plans which were adopted by Congress 23 May 1850. These schedules comprised: (1) Free inhabitants; (2) Slaves; (3) Mortality; (4) Products of agriculture; (5) Manufactures; (6) Social statistics. But the Freat advance which it made was in scheduling individuals instead of fami lies, thereby making full statistical comparison for the first time possible. The name, age, sex, color and place of birth of each person, free or slave, were now set down. Improvements in method were also made. The enumerators' dis tricts were to be not exceeding 20,000, if feas ible, on the basis of the last census; the returns were not to be compiled by the marshals, but at a central office in Washington created within the Department of the Interior. Its head or "supervising clerk,)) commonly called °superin tendent of the census," was Joseph C. G. Ken nedy of Pennsylvania; the latter part of the compilation was done under the charge of James D. B. De Bow of Louisiana. There were 45 districts, and the work cost $1,423,350.75.
The eighth census, that of 1860, was on much the same lines as the previous one. Mr. Kennedy was superintendent. Fortunately, the enumeration was finished before the war broke out. The principal change in schedules was to classify females as well as males by occupation. In the compilation, slaveholders and slaves were classified .by States and Territories for 1850 and 1860. There were 64 marshals and secretaries. Cost, $1,969,376.99.
The ninth census, that of 1870, was taken under changed conditions, some of which inter fered with its completeness and accuracy. The mass of enfranchised negroes in the South could no longer be counted on plantations by one authority, and were too ignorant to answer questions for themselves; and the Re construction conditions greatly impeded work. A new system had been carefully drawn up by Garfield, after consultation with Francis A. Walker, Edward Jarvis and other eminent statisticians, which would have made partial amends; but it was killed in the Senate. It formed, however, the basis of the next census. General Walker, then chief of the bureau of statistics in the Treasury Department, was made superintendent, and his ability did much to redeem the defective system. The heading °slaves° had of course to be dropped and the number of male citizens of the United States to be substituted. To satisfy the 15th Amend ment, a schedule was introduced of the num ber of such citizens whose right to vote was denied or abridged except for crime, etc.; but
it was valueless. The month of births and marriages within the year was added. In "col ored," Chinese and Indians were discriminated from negroes. In “illiterates," those unable to read were distinguished from those unable to write. In nativities, the place of parents' birth was added for the first time giving the number of native citizens of foreign parentage. A statement of public debts, State, municipal, etc., was included. Many changes were also made for clearness, or to remove objectionable quiries. The enumerators were put under civil service reform rules.
A strong effort was made to have an inter decennial census taken in 1875, but it was not successful, though Grant recommended it in two annual messages. But for that of 1880, the system desired for 1670 was passed, under Gar field's presidency, completely revolutionizing the census methods. The changes may be summa rized as follows: (1) The superintendent of the census and his chief enumerators, instead of being mere clerks of the Secretary of the Interior, became officials directly appointed by the President and confirmed by the Senate. (2) The marshals — court officers appointed for du ties unconnected with census work and often conflicting with it— were replaced by super visors of the census, not to exceed 150 in all, and each State or Territory to have at least one. The entire number were appointed and New York State had 11. Their assistant enumera tors were subject to veto by the census office. (3) The enumerators' districts were reduced to an estimated maximum of 4,000 instead of 20,000, these subdivisions being also subject to the disallowance of the census office. This increased the number of enumerators from about 6,400 in 1870 to 31,265 in 1880; and en abled not only a swifter completion of work, but much closer local and personal knowledge on the part of enumerators, as each must be a resident of his district. (4) The time of enu meration was reduced to one month, or two weeks in cities of over 10,000 inhabitants by the previous census; making results more nearly simultaneous. (5) Compensation was more closely adjusted to work. (6) Industrial sta tistics for 279 cities and large towns, and for the country at large in certain selected indus tries, were withdrawn from the regular enumer ators and given to certain special agents, who need not be residents. The enormous advan tage of this scarcely needs exposition. Not only did it allow the employment of abler experts, but manufacturers will often give information to distant strangers, to be filed far off, which they would not put in the hands of local residents. In this census also a large use was made of prior schedules,* left with parties before the enumerator called; they had been introduced in 1870 but little employed. This census was by far the best, speediest and most helpful of all to that date and marked another era. The scope was also extended. For the first time the statistics of defectives, crime and pauperism were thoroughly and properly classified; those of the factory system, of churches, museums, taxation, insurance, etc., entirely new, with better ones of schools and colleges, collected independently, were of great utility; and many improvements were made in other schedules. General Walker was the head of this census also and the author of most of its advances. Its cost was $5,790,678.40.