Note on Revisions: The following revisions were made to the Agricultural Workforce files in Oct. 2007 to correct some minor errors that were found in the underlying population statistics by John Brown. I am grateful to John for bringing these to my attention.

In 1860, the "free" population totals for Brazoria County (code #390) and Brazos County (code #410)in Texas, were switched from what was shown in the published census. Brazoria County’s free total was shown as 1713 in the data set, but should be 2033 and Brazos County's was shown as 2033 in the data set, but should be 1713. This correction also results in a change in the total population in each of the two counties, but had no effect on the workforce estimates because the population components by age, sex and free-slave status were correct.

In 1850 in the Darlington District (code #310) of South Carolina, the free and total population figures were too large by 1 (the correct figures are 6,789 and and 16,830, respectively. And, in the Union District (code #870) of South Carolina, the free and total population figures were too small by 1 (the correct figures are 9,460 and 19,852, respectively.

Also in 1850 the urban population of the town/city of Lafayette, Louisiana was incorrectly assigned to Lafayette Parish (code #550), when it should have been placed in Jefferson Parish (code #510). This correction alters the county total population, and because the size of the urban population has some influence on the estimation of the agricultural workforce, those figures have changed slightly as well.

Finally, in 1840 Maryland, there was a special act in 1841 which led to a re-count of the slave population in Montgomery County (code #290). There is a footnote on p. 30 of the Compendium volume published in 1841 to this effect. While the original 1840 census has a total of 5,135 slaves (as in the data set), the re-taken count totals 5,377 slaves (which is found in the Compendium volume). Although we do not know the details or circumstances leading to the re-count, we have assumed that the 5,377 is the more accurate count. This correction raises the county’s total population, and the agricultural workforce as well.

Tom Weiss, University of Kansas