New evidence based on census data indicates that output per agricultur
al worker grew faster between 1860 and 1870 than during any other deca
de of the nineteenth century. Although this evidence seems to support
the traditional view that the Civil War was a catalyst for an increasi
ngly productive agricultural sector, we contend that this apparent rob
ust performance results from a measurement problem that afflicts censu
s-based labor force series. An alternative estimate of labor force per
formance during the decade reveals the importance of increased labor i
nputs of women and children, in numbers, effort, and-especially-time.