You are here: Economic Programs Overview 
Manufacturing
Longitudinal Research Database
LONGITUDINAL RESEARCH DATABASE
PURPOSE
To provide a company-level database containing detailed statistics on research
and development activities; and support research on the issues of productivity,
profitability, and the uses of research and development. The United States Code,
Title 13, authorizes this data program. The National Science Foundation provides
significant funding for the program and related research.
The R&D database
information is confidential [Title 13 and Title 26, US Code]. All research is
conducted in the Center for Economic Studies (CES) by persons permanent and
specially sworn to uphold the confidentiality provisions ot Title 13 and Title
26.
COVERAGE
The database contains detailed company-level research and development information
compiled from the annual Industrial Research and Development survey
for survey years 1972 through 2001. Over this period of time, the scope of the
survey has been adjusted to reflect the changing nature of the U.S. economy.
In earlier years, the emphasis was on the manufacturing sector of the economy;
in more recent years, interest in the non-manufacturing sectors has increased.
Over the thirty-year period, the total sample for the survey size has varied considerably.
The total sample size has been fairly stable at approximately 25,000
companies since 1992. However, in any given year, the number of sampled companies that
actually conduct or sponsor R&D activities is generally in the 3,500 to
4,000 range. Due to the concentration of R&D activities among the larger
companies, most companies with significant R&D activities remain in the
sample for a number of years.
CONTENT
The database includes information on R&D spending and company
characteristics, and data collected under both mandatory and voluntary
survey response conditions. Mandatory data include total R&D,
Federal versus company financed R&D, domestic net sales, domestic
employment and R&D by State (starting in 2001). Voluntary data
are more detailed and include the number of scientists and engineers;
basic and applied research and development by Federal and company
funds; contracted-out, foreign, and budgeted research; Federal research
and development performed by principal government agency; R&D
by major type of expense and technology area; contracted-out R&D
by type of organization; Federal and company state R&D, energy
R&D, R&D performed in collaboration with others by type
of organization, and R&D by country.
FREQUENCY
The underlying survey data have been collected annually on a calendar
year basis since 1957; however, only data from 1972 forward is available on
the database. The R&D database is generally updated annually within 2 years
after the survey reference year.
METHODS
The sample design strategy has evolved over the years. The company
has been defined as both the sample unit and the data collection unit since
inception. Prior to 1992, a given sample would be used for a number of years
before replaced. The probability of selection was a direct function of total
company employment; companies with more than 500 employees were included with
certainty. Companies with less than 500 employees had probabilities varying
between .001 and .999. Sampled companies who reported that they did not engage
in any R&D activity were removed from the sample.
A new sample has been selected for each cycle of the
survey since the 1992 survey. This has improved the representativeness of the sample and has allowed
us to introduce sample design enhancements in response to changes in the survey
requirements or priorities. We continue to use a probability-proportionate-to-size
approach for assigning the individual company probabilities; but the specific
methodology has varied over the years.
The data is collected via a mail-out/mail-back process, with limited
electronic or telephone collection. Upon receipt the response data
is subjected to a series of edits designed to assure that the data
is both reasonable and consistent.
Periodically, a limited number of visits to key companies are conducted
to gain insight to the changing nature of R&D activities, to
better understand any reporting difficulty the companies may have,
and to determine the collectiblity of proposed new items.
R&D activity is a highly skewed activity. The largest 1,000
companies with R&D activity account for 90% of the total R&D
expenditures in the U.S. Thus, the database is primarily comprised
of larger companies.
The R&D database information is confidential and all research is conducted
by permanent and specially sworn Census Bureau employees. All current research
is done at the Center for Economic Studies (CES) in Suitland, Maryland.
PRODUCTS
The CES Discussion Paper series began in 1988, and periodically presents working papers based
on R&D research. Current titles include Using
Linked Census R&D-LRD Data to Analyze the Effect of R&D Investment on
Total Factor Productivity Growth”, and Academic Science, Industrial
R&D, and Growth of Inputs. The database is described in the title
A Guide to R&D at the Center for Economic Studies. Most papers
are revised and submitted for journal or commercial publications. Special tabulations
of aggregated data from the R&D database are prepared in response to specific
requests on a reimbursable basis.
USES
The National Science Foundation has sponsored research to measure and assess
the diverse effects of industrial R&D. Planned research will explore the
lagged and concurrent relationships between university-sponsored R&D, firm-sponsored
R&D, and firm productivity.
RELATED PROGRAMS
Economic Census
Annual
Survey of Manufactures
Current Industrial Reports
(Industrial Products)
Report of Organization
(Company Organization Survey)
Pollution
Abatement Costs and Expenditures Survey
Manufacturing Energy
Consumption Survey
You are here: Economic Programs Overview
Manufacturing Longitudinal Research Database
Last revised:
July 27, 2006
