Economic historians are wont to set the record straight. In part, this desire stems from the aesthetic value of seeing each jot and tittle in its proper place, and from the comfort found in believing that nothing is askew. We sleep a little better at night knowing that sound data are in the computer terminal. There is also a bit of the detective in each of us, and our risk aversion leads us to sleuth among wayward estimates rather than among hardened criminals. There is, of course, the more practical, and we suspect more valuable, purpose of assuring the users of historical time series that the data are accurate and consistent. The accuracy and consistency of any time series rest on the replicability of the original estimates and procedures, and on a masochistic streak which occasionally compels some of us to examine in detail the original estimates, replicating the various parts and rendering an assessment of the data. In some cases revisions are forthcoming and the record is set straight again-at least for a while. It has been almost 20 years since Stan Lebergott's estimates of the nineteenth-century workforce appeared in print and some initial revisions were suggested (Lebergott 1966; David 1967). These estimates

Thomas Weiss is professor of economics at the University of Kansas and a research associate of the National Bureau of Economic Research. The paper has benefited from the comments of Paul David, Stan Engerman, Robert Gallman, and the research assistance of Terri von Ende. I also wish to thank Stan Lebergott for patiently answering my questions and for giving me access to his worksheets. The work was financed in part by the University of Kansas General Research Fund and the National Science Foundation (No. SES 8308.569). This research is part of the NBER's program in Development ofthe American Economy. Any opinions expressed are mine and not those of the National Bureau of Economic Research.