The U.S. Census as an Historical Source

The Federal censuses are an invaluable tool for historians and other social scientists. For the Civil War and Reconstruction era alone, historians have used the enumerations to estimate voter turnout in elections, to investigate the ethnic, religious, and socio-economic patterns of party affiliation, and to gauge the relative strength of the two sections on the eve of the war. They have used the figures to assess the impact of the war and emancipation on our economy and society, and, as with this project, to reconstruct the life-stories of individuals and communities. Census pages allow historians to recreate neighborhoods and kin networks. The vast majority of this research has been carried out within the past twenty-five years, as computer technology has enabled scholars and students to use this data with increasing ease and methodological sophistication.

Despite their value, the census counts are nevertheless far from perfect. Like any historical source, they must be evaluated, their strengths noted, and their limitations recognized. In general, scholars agree that the nineteenth-century censuses tended to undercount the true population of the United States. This was due to several factors, including the protracted length of the enumeration (which during the antebellum period often took more than three months to complete) and the mobile nature of the American population. Underenumeration was a particular problem in big cities where crowded living conditions and irregular work hours made it harder to catch people at home. So, too, we know today that poorer individuals and recently arrived immigrants are less likely to understand or trust the intentions of the census-taker, whom they may associate with the government's proclivity to tax, draft, or arrest the poor. Even in rural districts, however, enumerators often found it difficult to reach people at home. While the Augusta and Franklin manuscript returns for 1860 are fairly complete, we know from newspaper articles that many of the problems faced by enumerators, including suspicion of the purposes of the census and the giving of inadequate or faulty information, were present in these counties as well.

The accuracy of the census counts also depended upon the conscientiousness of the enumerators, most of whom were untrained and poorly supervised. In the nineteenth century, these posts were often distributed as political patronage, with little consideration given to the fitness of specific individuals for the job. Hence, the commitment of individual enumerators varied enormously. There is no question that mistakes in the censuses abound, particularly in the earlier years. Even if the data is presumed to be sufficiently correct for the question at hand, historians must take care not to use the numbers in erroneous or misleading ways.

Complete federal census data for Augusta County, Franklin County, and other aggregated units, from 1820 through 1860, can be found under Searching the Aggregated U.S. Censuses, 1820-1860. Since these data are composed primarily of absolute numbers without respect to population, be sure to use them with care! Help in calculating percentages and other simple statistics is provided by clicking on Strategies at the bottom of each page.