Restricted data sets available through FSRDC are becoming central in economics research. The range of these studies is incredibly wide; the brief overview below barely scratches the surface. However, it highlights the key feature that individual observations on, e.g., workers, firms, etc., can be linked across data sets available at FSRDC and to external data sources that contain relevant individual identifiers. This drastically expands the scope of the analysis that was previously possible.
While there are various limited public sources of microdata on individuals, there is virtually no public use microdata on firms and establishments in the US. Such data are only available through FSRDC. Moreover, available data not only cover all sectors of the economy but also allow to link individual observations across time. At the foundation of such analysis is the Longitudinal Business Database (LBD) which covers 8.5 million establishments over 1976-2011. Establishments have a permanent ID and can be linked across time and aggregated to firms using firm ID. Crucially, establishment ID can be linked to other data sets.
For example, the Bureau of Customs and Border Protection collects information on virtually the universe of export and import transactions, containing value, date, country of origin or destination, quantity, detailed product code, and other variables. Importantly, each transaction is linked to individual importing or exporting establishment ID which is the same as in the LBD allowing to study the dynamics of firms engaged (at some point of their existence) in international trade.
Moreover, while LBD provides only basic information on establishments, such as their geographic location, industry affiliation, total annual payroll and employment, much more information is available by linking LBD to other Census Bureau censuses and surveys, which include among many others, Quarterly Financial Reports, R&D and Innovation Surveys, Capital Expenditure Surveys, quarterly Manufacturing Plant Capacity Utilization Surveys, etc.
Another aspect of data linkage is matching of the data on individual establishments and their workers. These data come from the Longitudinal Employer-Household Dynamics (LEHD) program and represent the only matched employer-employee dataset in the US. As the name implies, the data identify each US workers’ employer in any given quarter and reveal a complete set of co-workers of every worker. These data are currently at the cutting edge of research in labor and macro economics. For example, recent research has identified how these data can be used to reveal (unobserved to economists) productivity rankings of all workers and all firms in the economy and the complementarities between workers and firms in production, which allows the study of the efficiency of the observed allocation of workers to firms and productivity-enhancing changes to this assignment.
LEHD data can be further linked to LBD and other establishment censuses and surveys to obtain a rich set of establishment characteristics. Individual workers in LEHD can also be linked to Decennial Censuses and other surveys, such as American Community Survey, Current Population Survey, Survey of Income and Program Participation, etc. to obtain rich set of demographic and other characteristics of individual workers. Finally, workers and firms or establishments in LEHD can be linked to outside sources of data. For example, recent research has linked LEHD with credit reports on individual workers and business owners to study the role of credit constraints in business formation and the relationship between credit access and job flows.
Selected Current Penn Proposals in Progress
Aging and Public Policy: Implications for Health, Innovation and Welfare
- Harold Cole, Department of Economics, University of Pennsylvania,
- Soojin Kim, Purdue University,
- Dirk Krueger, Department of Economics, University of Pennsylvania,
- Alexander Ludwig, Goethe University of Frankfurt
Localized Knowledge Spillovers: Evidence from the Agglomeration of American R&D Labs and Patent Data
- Geraldo Carlino, Federal Reserve Bank of Philadelphia,
- Bob Hunt, Federal Reserve Bank of Philadelphia,
- Tony Smith, School of Engineering and Applied Science, University of Pennsylvania,
- Kristy Buzard, Department of Economics, Syracuse University.
The Interaction of Labor and Housing Markets
- Fernando V. Ferreira, Real Estate and Business Economics, Wharton School, University of Pennsylvania
Worker-Firm Sorting in the U.S. Labor Market
- Iourii Manovskii, Department of Economics, University of Pennsylvania
Open-Source Federal Budget Model
- Kent Smetters, Professor, the Wharton School, University of Pennsylvania
Other FSRDC Economics Research Examples
Every FSRDC project using Census data submits a working paper to the Discussion series, downloadable at: https://ideas.repec.org/s/cen/wpaper.html