Foreign Direct Investment

FDI Questions and Answers

1. Will Foreign Direct Investment data be regularly produced?

BLS is releasing these data on a research basis. Future releases will be based upon continued demand for the data and availability of resources.

2. When will the next updated version of these data be released?

There is currently no established date for an update.

3. Why did BLS choose to publish data for 2012?

BEA’s most comprehensive survey of foreign direct investment, in terms of both coverage of companies and the amount of information collected, is performed every five years. The 2012 Benchmark Survey of Foreign Direct Investment in the United States is the most recent benchmark year available. The 2017 data will be available in late 2019.

4. How can a foreign-owned company have an ultimate beneficial owner from the United States?

A U.S. company is foreign owned if a foreign entity directly or indirectly holds an ownership share of 10 percent or greater. (Ten percent ownership is the threshold between direct investment and portfolio investment.) In many cases, the foreign entity that owns a U.S. company is owned in turn by another entity or another chain of entities. The entity at the top of the ownership chain is the ultimate beneficial owner (UBO) of the foreign-owned U.S. company. In most cases, the entity at the top of the ownership chain is a foreign entity, but sometimes it is a U.S. entity.

5. Are there earlier produced BLS data on employment and wages in companies with foreign ownership? Where can I find these data?

Beginning with the enactment of the Foreign Direct Investment and International Financial Data Improvements Act of 1990, BLS and BEA collaborated to combine BEA data on foreign-owned businesses with BLS employment data until the funds dedicated to the project were eliminated. That project, which focused on the composition of domestic employment related to direct investments in the United States by foreign firms, produced the tabulations “Employment and Wages in Foreign-Owned Businesses in the United States” for 1989 through 1992 and “Occupations in Foreign-Owned Manufacturing Establishments in the United States” for 1989. These data can be accessed through archived BLS news releases:

6. Why are the research data from the link of the 2012 Benchmark Survey of Foreign Direct Investment in the United States with BLS’s Quarterly Census of Wages and Employment (QCEW) and Occupational Employment Statistics (OES) different from BEA’s statistics on the activities of U.S. affiliates of foreign multinational enterprises (MNEs)?

The research data are not directly produced from data collected on the 2012 Benchmark Survey. They are from either the QCEW or the OES and cover establishments that have been matched to U.S. affiliates in the Benchmark Survey. Differences between the research data and the statistics on the activities of U.S. affiliates of foreign MNEs arise from several factors:

  1. Differences in scope. Statistics on the activities of U.S. affiliates include employment for U.S. territories, which are not covered by the QCEW and which were not included in the OES data used for the research data. The statistics also include employment for certain worker categories that are excluded in the QCEW or OES. For example, the QCEW excludes railroad workers covered by the railroad employment insurance system and the OES excludes most of the agricultural sector.
  2. Differences in timing. Statistics on the activities of U.S. affiliates typically reflect employment at the end of a U.S. affiliate’s fiscal year ending in 2012. QCEW estimates in the research data are an average of quarterly reports over the 2012 calendar year. OES estimates in the research data are derived from a sample survey collected from 2011 to 2013 in 6 survey panels (in May and November of each year).
  3. Incomplete or inaccurate matching. Despite the best efforts of BLS and BEA, some establishments may not have been matched to a U.S. affiliate even though they were foreign owned, and some establishments may have been matched to a U.S. affiliate even though they were not foreign owned. For more information, see "What should I know about the process of matching U.S. affiliates from BEA’s 2012 Benchmark Survey of Foreign Direct Investment in the United States with establishments from the BLS’s Quarterly Census of Employment and Wages (QCEW)?"
  4. Reporting or imputation errors. Companies may inadvertently misreport employment on the Benchmark Survey, the QCEW, or the OES. Companies may fail to report on one or more of these, which necessitates imputation by BEA or BLS, and these imputations by one Bureau may not match imputations by, or data reported to, the other Bureau. For the OES research data, imputations are made for units that did not respond.

For more information on the linked dataset and to access the data, see our data page.

7. Which data should I use for understanding 2012 employment at U.S. affiliates: BEA’s statistics on the activities of U.S. affiliates produced from the Benchmark Survey or data from the research dataset?

For most purposes, statistics on the activities of U.S. affiliates of foreign multinational enterprises (MNEs) from the 2012 Benchmark Survey of Foreign Direct Investment in the United States are the primary resource for understanding employment at U.S. businesses with foreign ownership in 2012. This is particularly the case for comparing employment with other U.S. affiliate data items such as property, plant, and equipment expenditures or sales. The Benchmark Survey collects employment data together with these and other data items, ensuring consistency in scope and timing. Use of the activities statistics also avoids errors arising from incomplete or inaccurate matching of BEA affiliates with BLS establishments. Finally, for understanding employment at foreign-controlled U.S. affiliates (those with more than 50 percent ownership by foreign direct investors, also referred to as “majority-owned” U.S. affiliates), the statistics on the activities of U.S. affiliates are the sole source of information; the research data cover employment at all U.S. affiliates—both foreign-controlled U.S. affiliates and affiliates with between 10 and 50 percent ownership by foreign direct investors—but does not provide a breakout for foreign-controlled affiliates versus other foreign-owned U.S. affiliates.

The research dataset offers information for understanding employment at U.S. affiliates at very detailed levels. The research dataset provides information for many counties and metropolitan statistical areas, whereas the greatest level of geographic detail provided in Benchmark Survey statistics is U.S. states. In addition, the research data provide finer industry detail than the activities of U.S. affiliates of foreign MNEs statistics. The industry detail in the research data, particularly for wages, may also more accurately reflect the full range of industries in which U.S. affiliates operate, as the research data classify each establishment into its primary industry, while the activities statistics’ wage estimates assign a single industry to each U.S. affiliate, which may consist of establishments in several different industries. (For employment, the activities statistics are also alternatively classified by industry of sales, which more closely approximates industry of establishment than does industry of affiliate.) Finally, information on occupations is only available in the research dataset.

For more information on the linked dataset and to access the data, see our data page.

8. What should I know about the process of matching U.S. affiliates from BEA’s 2012 Benchmark Survey of Foreign Direct Investment in the United States with establishments from the BLS’s Quarterly Census of Employment and Wages (QCEW)?

Establishments with foreign direct investment were identified in the BLS QCEW longitudinal database through an initial match to U.S. affiliates using an identification number common to both agencies. This mutual identifier, the employer identification number (EIN), is a unique nine-digit number issued by the Internal Revenue Service to identify a business entity. However, employers may have more than one EIN, and they may, for example, use different EINs for different subsidiary companies, or different EINs for different purposes. Matching through EINs was automated, and the quality of the resulting links was evaluated—primarily by comparing the sum of employment in matched QCEW establishments to employment at the corresponding U.S. affiliate in the Benchmark Survey data. A match for which the corresponding employment values were within 20 percent of each other was considered to be adequate. Further manual matching efforts were undertaken for matches deemed to be inadequate. Matching via EIN could result in either overmatching or undermatching in some cases. It could find QCEW employment higher than the associated Benchmark Survey employment if, for example, two different establishments used the same EIN but only one is foreign owned; this was particularly common with investment firms and the companies they owned. Conversely, matching via EIN could find QCEW employment lower than the associated Benchmark Survey employment because the Benchmark Survey does not always collect every EIN used by the U.S. affiliate. Manual review resolved some undermatches or overmatches from automated EIN matching, particularly for larger U.S. affiliates, and confirmed, in certain cases, legitimate EIN matches with employment differences of more than 20 percent. Nonetheless, some inaccuracy likely remains.

For more information on the linked dataset and to access the data, see our data page.

Last Modified: February 4, 2019