Search this site:

4 Data Sources

While most of the information for this Chapter is derived from conventional literature sources, on-line databases were utilized to estimate workforce and productivity. The primary source of county-level timber product output (TPO) data was the Forest Inventory and Analysis (FIA) TPO Database Retrieval System (Anonymous 2000). This database contains information about roundwood products harvested in each county for calendar year 1996, by species and product class.

There are several sources of logging employment data. County-level data were obtained from civilian labor force data of the 1990 decennial census (U.S. Census Bureau 2000a), the most recent available sample of self-reported employment status. A sample of 1990 Census respondents described their industry and occupation. Based on this information, people were assigned to standard occupational and industry codes. Total logging employment was assumed to consist of both occupations 496 (Timber cutting and logging) and 494 (Supervisors, forestry and logging). State-level logging employment data for the period 1997-1999 were derived from the Covered Employment and Wages Program (ES-202) of the Bureau of Labor Statistics (Bureau of Labor Statistics 2000a). The ES-202 data are a 100 percent report for all establishments covered by unemployment compensation insurance. Older State-level workforce data were compiled from the Annual Survey of Manufactures (ASM) (U.S. Census Bureau 2000b). ASM data are collected through a mail survey of a sample of establishments. Both the ES-202 and ASM were queried for total State-level employment in Standard Industrial Classification (SIC) Code 241, Logging. Some years of the ASM data are missing for Kentucky, Tennessee, and Oklahoma.

The 1990 Census provides a snapshot of logging employment at the county level. Because it is based on self-reported occupation, it may provide a more accurate measure of workforce in an industry with many small firms and self-employed workers. However, it is also subject to errors in classification, and some nonloggers are likely included in the 494/496 occupational codes. The annual data from the ES-202 and ASM surveys provide an employment time series, but likely underestimate the logging workforce because they are based on a sampling of establishments. A comparison of the 1990 workforce at the county and state-level highlights the possible disparity. The ASM data estimate a total southern (less KY,TN,OK) logging workforce of 36,000. In comparison, the decennial census estimates a total of 44,066. Most States are within several hundred workers. Texas, Virginia, and Mississippi, however, account for 6,000 of the 8,066 difference in workforce estimates. For this report, the decennial census data were considered a reasonable estimate of workforce and the manufacturer survey data were used to model trends over time.