African American Census Schedules
From about 1830 on, northern cities increasingly felt the need to monitor African Americans who were moving from the South seeking freedom and work. In 1863, in the midst of the Civil War, Ohio called for the number and names of African Americans who had immigrated to Ohio from other states since 1 March 1861, their current township of residence, and their state of origin. Thirteen counties in southeastern Ohio submitted schedules. Hamilton County refused because the numbers were too great and its staff too limited.
Household censuses of Philadelphia's African American population were taken in 1838 and 1856 by the Pennsylvania Abolition Society and in 1847 by the Society of Friends. In addition to the variables listed in the federal census, the records of 11,600 households contain information describing membership in church, beneficial, and temperance societies; income, education level, and school attendance; house, ground, and water rent; how freedom was acquired; and the amount of property brought to Pennsylvania. These superb records constitute the most detailed information we have describing any population group in the mid-nineteenth century; they are being computer-processed as part of an urban-immigrant study of African Americans in Philadelphia conducted by Temple University.
The National Archives has issued a separate list of 'Free Black Heads of Families in the First Census of the U.S. 1790' as Special List 34. This compilation by Debra L. Newman is available free of charge upon request from the National Archives. An expanded version for New York is Free Black Heads of Households in the New York State Federal Census 1790'1830, compiled by Alice Eichholz and James M. Rose.<ref>Alice Eichholz and James M. Rose, comps., Free Black Heads of Households in the New York State Federal Census 1790'1830, Gale Genealogy and Local History Series, vol. 14 (Detroit: Gale Research Co., 1981).</ref>
NARA has also published List of Selected African Americans from the 1890 and 1900 Federal Population censuses of Delaware and Related Census Publications 'Agriculture in the State of Delaware' (1901) and 'Negroes in the United States' (1904), which reproduces lists of selected African Americans from the 1890 and 1900 censuses of Delaware as well as other related Bureau of the Census publications. You can find it on the NARA website.
African Americans in the Federal Census
It has been widely noted that African Americans were enumerated as all other U.S. residents from 1870 (the first census year following the Civil War and emancipation) onward. Prior to 1870, however, the situation was far different. Although free African Americans were enumerated by name in 1850 and 1860, slaves were consigned to special, far less informative, schedules in which they were listed anonymously under the names of their owners. The only personal information provided was usually that of age, gender, and racial identity (either black or mulatto). As in the free schedules, there was a column in which certain physical or mental infirmities could be noted. In some instances the census takers noted an occupation, usually carpenter or blacksmith, in this column. Slaves aged 100 years or more were given special treatment; their names were noted, and sometimes a short biographical sketch was included. In at least one instance, that of 1860 Hampshire County, Virginia, the names of all slaves were included on the schedules, but this happy exception may be the only instance when the instructions were not followed.
Sometimes the listings for large slaveholdings appear to take the form of family groupings, but in most cases slaves are listed from eldest to youngest with no apparent effort to portray family structure. In any event, the slave schedules themselves almost never provide conclusive evidence for the presence of a specific slave in the household or plantation of a particular slaveowner. At best, a census slave schedule can provide supporting evidence for a hypothesis derived from other sources. Prior to 1850 there were no special slave schedules for the manuscript census, as slave data was recorded as part of the general population schedules. In these, only the heads of household were enumerated by name.
In the absence of any contradictory information, it might be assumed that a family of freed people enumerated in the 1870 census was living not far from its last owner, whose surname they also bore. There would, of course, be reasons to dispute both assumptions. (Knowledge of the Civil War history of a locality could come into play here; for example, such relative stability would not have existed in a Georgia county that was in the path of Sherman's march to the sea.) Even so, this assumption represents one of the more obvious exploratory lines of research, especially in the absence of any other options. The first step in testing the hypothesis would be to search for slaveowners of the same surname in the 1860 slave schedules of the county in which the African American family resided in 1870.
Starting in 1850, another supplemental schedule, the mortality schedule, listed all deaths within a year before the regular census enumeration. The deaths of blacks and mulattoes, both free and slave, are recorded in them, even though their names have not been included in many of the indexes to these schedules. The deaths of slaves were generally enumerated in four fashions: unnamed (as in the slave schedules), but perhaps with the owner identified; by first name only; by first name and surname; and by first name with the owner noted.