Linking firms with establishments in BLS microdata
An examination of the Employer Identification Numbers (EINs) of a sample of large firms that use multiple EINs reveals that only a small percentage of their establishments and employment can be identified with the one EIN that each such firm uses in its filing of Securities and Exchange Commission Form 10-K; approaches are suggested to overcome this limitation.
The Bureau of Labor Statistics (BLS, the Bureau) collects data from employers about their establishments. For certain applications, however, researchers inside and outside the Bureau need data on firms. For example, in an earlier Monthly Labor Review article, Elizabeth Handwerker, Mina Kim, and Lowell Mason attempted to find all of the establishments associated with the 500 largest multinational manufacturing firms identified in surveys conducted by the Bureau of Economic Analysis (BEA).1 Other researchers have suggested merging BLS microdata with additional datasets containing information about firms.2 This article (1) gives an overview of the complex relationship between firms, on the one hand, and their establishments and establishment identifiers, on the other, and (2) outlines the efforts involved in linking establishment data into firms.
The backbone of all employer microdata at the Bureau is the Quarterly Census of Employment and Wages (QCEW). Covering approximately 9 million establishments nationwide and 98 percent of U.S. employment, this dataset contains quarterly records of all U.S. business establishments subject to state Unemployment Insurance (UI) laws.3 The records of the QCEW include monthly employment and quarterly total payroll data, based on the quarterly contribution reports employers submit to the state agencies responsible for administering UI programs. Each establishment in the QCEW is an economic unit, such as a farm, mine, factory, or store that produces goods or provides services. Establishments typically have a single physical location and are engaged in one type of economic activity.
In recent years, several researchers have expressed interest in merging corporate datasets compiled from firms’ mandatory filings with the Securities and Exchange Commission (SEC) with QCEW data, using firms’ federal Employer Identification Numbers (EINs) as the identifier for linking firm data to the establishment data of the QCEW. However, there is no simple way to use EINs to find, for a given firm, all of that firm’s establishments in the confidential microdata of the QCEW. Although every establishment in the QCEW is associated with both a federal EIN and a state UI account number, businesses may use one EIN for the UI tax system and other, different EINs for other tax systems. Put another way, both EINs and UI account numbers define businesses for tax purposes, but a firm may have more than one EIN and more than one UI account number. Thus, firms may use one EIN in filings with the SEC and a different EIN (or set of EINs) in reporting to the UI system. Also, firms that span multiple states will have a different UI account in each state, and large, complex firms may use numerous EINs across many states.
The BLS Business Employment Dynamics program publishes estimates by firm size, based on QCEW data. These estimates, however, are calculated at the EIN level. In other words, there are no true firm identifiers, other than EINs and UI accounts, in the QCEW.
1 Elizabeth Weber Handwerker, Mina M. Kim, and Lowell Mason, “Domestic employment in U.S.-based multinational companies,” Monthly Labor Review, October 2011, pp. 3–15.
2 On a limited basis, the Bureau allows eligible researchers to access confidential data for purposes of conducting valid statistical analyses. (For more information, see “BLS information: researcher access to confidential data for purposes of conducting valid statistical analyses. (For more information, see “BLS information researcher access to confidential data files at the BLS” (U.S. Bureau of Labor Statistics, Aug. 22, 2011), http://www.bls.gov/bls/blsresda.htm.)
3 Employment data exclude members of the Armed Forces, the self-employed, proprietors, domestic workers, unpaid family workers, and railroad workers covered by the railroad Unemployment Insurance system.