The Census of Construction Industries (CCI) is conducted every five years as part of the quinquennial Economic Census. The CCI covers all establishments with payroll that are engaged primarily in contract construction or construction on their own account for sale as defined in the Standard Industrial Classification Manual. The CCI is a partial census including all multi-establishments and all establishments with payroll above $480,000, one out of every five establishments with payroll between $480,000 and $120,000 and one out of eight remaining establishments. The resulting database contains for each year approximately 200,000 establishments in the building construction, heavy construction and special trade construction industrial classifications. This article compares the content, survey procedures and sample response of the 1982, 1987 and 1992 Censuses of Construction. In addition, it describes the Census Bureau's Economic Census Program, which includes the Census of Construction Industries.
The recent availability of establishment level microdata at the Census Bureau has spawned a renewal in the study of industry dynamics. The bulk of this research has utilized the Census Bureau's Longitudinal Research Database (LRD). In response to many of the insights and observations that have evolved from this analysis, the Census Bureau is currently encouraging the expansion of the LRD, which contains only manufacturing establishments, into a database containing all nonfarm business sectors. This new database, the Longitudinal Business Database (LBD), will be built from the various economic census data files. One of the first sectors outside of manufacturing to receive attention is construction.
In 1992, the construction industry had over 573,300 establishments with payroll, with over 4,674,000 workers, 3,596,000 of which were construction workers. Total value of construction work performed in 1992 was $529 billion. By comparison, the manufacturing sector in 1992 was made up of approximately 370,000 establishments, with close to 19 million workers, about 12 million of which were production workers. The manufacturing sector had a total value of shipments of over $3 trillion in 1992. This article overviews the contents, survey procedures and survey response of the 1982, 1987 and 1992 Censuses of Construction Industries, which currently constitute the Census of Construction Industries Database. Its primary purpose is to provide users with a comparison of the datasets across years, outline some of the strengths and weaknesses of the data, and allow them to explore the database's potential to answer relevant research questions.
The next section introduces the Census Bureau's Economic Census program, which includes the Census of Construction Industries (CCI). The following sections provide detailed information on the CCI, discuss the differences between the 1982, 1987 and 1992 versions of the CCI and summarize the procedures for linking establishments across the three years. Next, some of the strengths and weaknesses of the CCI are highlighted. The final section outlines how the reader may gain access to the CCI.
The Economic Census Program
The Census Bureau has taken an Economic Census every five years, in years ending with 2 or 7, since 1967. In addition to the construction sector, the economic census also covers the following areas: retail trade, wholesale trade, services industries, financial, insurance and real estate industries, transportation and communications, manufacturers and mineral industries. Special programs also cover enterprise statistics, and minority and women-owned businesses.
Since 1967, administrative records have been substituted for the surveying of very small firms and nonemployer establishments. In most industry sectors, a questionnaire is mailed to all establishments above a chosen size (either in terms of payroll or employees), while administrative data are used for all establishments below that level. …