Federal Facilities Research and Development (FFRD) Survey 2025

P urpose. The Federal Facilities Research and Development (FFRD) Survey collects information on research and experimental development (R&D) expenditures and personnel for all federally owned and operated facilities in the United States. This survey is the unique source of federal performer-reported data on R&D expenditures and R&D personnel. In conjunction with performer-reported data for the remaining sectors of the economy, FFRD Survey data can be used to estimate total national R&D performance and other critical statistics that are increasingly important in the analysis of global R&D competitiveness.

Data c oll ec tion authorit y . The information is solicited under the authority of the National Science Foundation Act of 1950, as amended, and the America COMPETES Reauthorization Act of 2010.

Survey contractor . ICF.

Survey sponsor . The FFRD Survey is sponsored by the National Center for Science and Engineering Statistics (NCSES) within the U.S. National Science Foundation.

Frequency . Annual.

Initial survey year . FY 2022.

Reference period . FY 2025.

Response unit . Establishment.

Sample or census . Census.

Population size . The population consists of 443 research-performing federal facilities within 36 federal agencies.

Sample size . Not applicable; the survey is a census of all known eligible federal facilities that conduct R&D.

Target population . The population for the FY 2025 FFRD Survey consisted of federally owned and operated facilities in the United States that performed R&D in FY 2025, excluding facilities of the Central Intelligence Agency (due to the classified nature of their work). Federally funded research and development centers (FFRDCs) are surveyed separately in the FFRDC R&D Survey, and University Affiliated Research Centers (federally funded but not federally owned or operated) are surveyed as part of the Higher Education Research and Development Survey.

Sampl ing frame . The facilities for the FY 2025 FFRD Survey were originally identified from the list of federal agencies with R&D obligations based on the Survey of Federal Funds for R&D and the Federal Laboratory Consortium list of laboratories. A facility is defined as a unit within an agency that is responsible for performing R&D, generally with its own distinct budget and leadership. Since each agency has a different organizational structure, this unit of measurement may be a division, branch, center, laboratory, or other entity and may span multiple locations. For FY 2025, the frame was updated by (1) reviewing records ineligible during the FY 2024 survey and determining which should remain ineligible and which should be included in the FY 2025 survey, (2) adding records having federal R&D obligations as reported in the Survey of Federal Funds for R&D, and (3) adding, removing, and adjusting reporting units based on communications with agency contacts. The total number of reporting units for the FY 2025 FFRD Survey is 289, which represents the 443 research-performing federal facilities.

Sample design.Not applicable.

Data collection. Data collection began with an email to each agency to verify the reporting units and to confirm or identify the name, phone number, and email address of each facility-level respondent. Table A-1 displays a list of facilities within each agency and notes which are reporting units for the FY 2025 FFRD Survey. When the reporting units are totaled, there are 289 total units across 36 agencies. There were 10 agency-level responses and 267 responses at a level lower than the agency-level for a total of 277 responses. Eight reporting units were determined ineligible, including one entire agency.

Respondents could choose to submit a questionnaire downloaded from the Web or sent via email (i.e., PDF or Excel format) or use a Web-based data collection system to respond to the survey. Questionnaires were carefully examined for completeness upon receipt. Respondents were sent personalized email messages asking them to provide any necessary revisions before the final processing and tabulation of data. These email messages included a link to the FFRD Survey Web-based collection system, allowing respondents to view and correct their data online. Respondents were also offered the opportunity to explain their data or provide corrections via email, through PDF, or through Excel and have their data updated by survey staff.

Data collection began on 21 January 2026, and the requested due date for data submissions was 31 March 2026. Data collection was extended until most surveyed agencies provided at least partial data with the last completed survey response verified in April 2026.

Mode . Respondents could choose to submit a questionnaire via the Web-based data collection system or through fillable PDFs or Excel files that were then imported to the Web-based data collection system. All edit and trend checks were accomplished through the Web-based system. Twelve reporting units submitted data using the fillable PDF, 151 reporting units submitted using an Excel version of the survey, and 114 submitted using the Web-based data collection system.

Response rates . By the survey's closing date in April 2026, forms had been received from 277 reporting units out of an eligible population of 289, a response rate of 95.8%. Table A-2 displays a detailed breakdown of response rates by Department of Defense (DOD) and non-DOD agencies, and table A-3 displays a breakdown of response rates for each survey question. Questions 9 and 10 had much lower response rates for non-DOD agencies and were the only questions with response rates less than 100% for DOD agencies. Questions 9 and 10 asked facilities to report headcounts for R&D personnel by function and job category (Question 9) and federal full-time equivalents (FTEs) by function (Question 10).

Data editing . Data inconsistencies or blank fields in the FFRD Survey were flagged automatically by the Web-based data collection system so that respondents could not submit their final data until all required fields were completed without errors. Respondents were contacted and asked to resolve possible self-reporting issues themselves. Questionnaires were carefully examined by survey staff upon receipt. Reviews focused on unexplained missing data, expenditures that significantly differed from the FY 2024 FFRD Survey, expenditures that significantly differed from the intramural obligations from the Survey of Federal Funds for R&D, and other data anomalies. If additional explanations or data revisions were needed, respondents were sent personalized email messages asking them to provide any necessary revisions before the final processing and tabulation of data. For any follow-up questions that went unanswered, NCSES was consulted before the data were either accepted without changes or adjusted based on information from other questions or previous contacts with the respondent.

Imputation . Instances of missing data occurred when a reporting unit did not respond to the survey (i.e., unit nonresponse) or when the reporting unit responded but did not answer certain survey questions (i.e., item nonresponse). After removing 8 ineligible reporting units, 289 total records remained, with 277 responses and 12 nonresponses. Among the respondents, total R&D expenditures were completely observed (from Questions 1, 2, or 5 totals). The pattern of missing data showed that 67% of reporting units have complete data across all variables selected for imputation. All personnel items are missing for 9% of reporting units. Another 21% of reporting units are missing some combination of personnel detail items or expenditure detail items. Respondents had no missing detail items for expenditure source, type, or field.

Imputed values are not displayed at the reporting-unit level; these are replaced with "NA" in the tables. However, the imputed values are included in the agency, department, and overall totals.

The imputation process for the FY 2025 FFRD Survey is a combination of multiple imputation models, and deterministic proportional allocation were used for item totals and item details respectively. Auxiliary information from the FY 2024 FFRD Survey was used in modeling and calculation of proportions where appropriate. The imputation includes two main steps:

Step 1. Key totals model.

Step 1a. Key totals estimation.

First, key totals were estimated using multivariate imputation by chained equations (MICE). The model included separate intercepts for each agency and log-transformed R&D totals. Individual models were evaluated for outliers, and no reporting units met the thresholds for exclusion. Multiple imputation was used to create multiple sets (m = 20) of plausible values for the missing values. After imputing total onsite R&D expenditures, the imputed value was assigned to other fields representing the same value including total type of R&D-type expenditures (Question 5), total field of R&D expenditures (Question 8), and within facility costs (subset of Question 2).

Step 1b. Expenditures details proportion allocation.

Next, the intramural total R&D expenditures were allocated to the subcategories of type of R&D, source of funding and federal agency source of funding, and field of R&D. Type of R&D and field of R&D were imputed for nonresponding reporting units based on the percent distributions using the following information in order:

1. FY 2024 FFRD Survey observed distribution for the reporting unit,
2. Responding FY 2025 reporting units within the same agency aggregated to the agency level,
3. Responding FY 2025 reporting units within the same agency aggregated to the department level, and
4. Responding FY 2025 reporting units within the same agency aggregated to the DOD or non-DOD level.

All imputed R&D amounts were assigned as a federal source of R&D. This follows the same procedure in FY 2022 and FY 2024 based on the majority of R&D coming from federal sources for each type (98% basic research, 97% applied research, and 99% experimental development).

Step 1c. Total expenditures and cost ratios.

After the MICE imputation model, the total R&D expenditures value was assigned to the cost type total (Question 2), since these two variables represent the same amount. Missing R&D expenditures performed outside facilities (subset of Question 2) were calculated as the difference between total R&D expenditures and total R&D within facilities.

For DOD agencies, when some reporting units within the agency responded (partial agency response), total research, development, testing, and evaluation (RDT&E) (Question 1) for nonresponding DOD reporting units was imputed based on a ratio adjustment using the ratio of total expenditures at the agency level and applying it to the imputed R&D total for the reporting units. When no reporting units within the agency responded to the FY 2025 FFRD Survey, the ratio adjustment was based on the department level from the FY 2025 FFRD Survey. Finally, operational system development expenditures were calculated as the difference between total RDT&E and total R&D.

Step 1d. Personnel details proportional allocation and total.

Lastly, personnel (Questions 9 and 10) totals were allocated to function (i.e., researcher, technician, support) based on the percentage distribution using the following information in order:

1. FY 2024 FFRD Survey observed distribution for the reporting units,
2. FY 2025 responding reporting units within the same agency aggregated to the agency level,
3. FY 2025 responding reporting units within the same department aggregated to the department level, and
4. FY 2025 responding reporting units within the same DOD or non-DOD category aggregated to the respective DOD or non-DOD level.

The imputed job counts and FTE for federal employees, contractors, and other employment status for each role were summed to calculate the imputed total job count.

Step 2. Mean aggregation.

The imputation process produced 20 multiply imputed datasets. These datasets were combined by taking the mean across imputations to yield a single set of reporting unit-level estimates for the final analytic file. For each variable, imputation flags were retained to indicate whether a reporting unit-level value was imputed. Quality control checks were conducted to ensure consistency between job counts and FTE measures. In cases where the imputed FTE exceeded the corresponding imputed job count, the FTE value was capped at the job count.

Weighting . Survey data were not weighted.

Sampling error . Because the FY 2025 survey was distributed to all eligible agencies performing R&D, there was no sampling error.

Coverage error. Under the total survey error framework, coverage error describes the difference between reporting units and units in the target population that the frame was developed to reach. The agencies included in the frame are those reporting nonzero intramural R&D obligations on the Survey of Federal Funds for R&D. It is highly unlikely that an agency would be performing R&D but not receiving any R&D obligations; therefore, coverage error should be close to zero.

Nonresponse error. Nonresponse error refers to the differences in key estimates between units in the frame that were selected for data collection and those that responded. Of the 289 eligible reporting units, 277 responded for a response rate of 95.8%. For unit nonresponse, multiple follow-ups were conducted with nonresponding reporting units, and multiple contact and data collection modes were used (i.e., phone and email) to mitigate nonresponse error. Missing data for nonresponding reporting units were imputed using prediction models to reduce potential nonresponse bias.

The item response rate ranged from 91.2% to 100% for the DOD agencies and from 81.5% to 100% for the non-DOD agencies. For item nonresponse, reporting units were encouraged to report estimates of expenditures when actual dollar amounts could not be provided. This approach reduces item nonresponse error risk but may introduce measurement error. Imputation was conducted to help mitigate item nonresponse error.

Measurement error . The FFRD Survey contained various ways for respondents to explain their answers. This included comment boxes for every question with instructions to provide additional information about how they calculated their response and any assumptions they made. Respondents were also prompted to enter comments if they indicated any data were unavailable. Additionally, questions requiring detailed reporting of personnel also included a checkbox to allow the respondent to indicate if they were able to provide the total number but were unable to provide details. The largest risk of measurement error is likely respondents' interpretation of the definition of R&D activities and variations in record-keeping procedures used by respondents to answer the survey questions. In addition, other known measurement problems are known to exist in the data, as shown below.

• R&D expenditures. The survey included a question asking respondents to confirm if they were reporting expenditures. Not all reporting units were able to report expenditure data and instead reported obligations. Thirty-seven reporting units confirmed they could not report expenditures.
• Type of R&D. Some reporting units said that these amounts were difficult to report either because they could not determine how best to allocate expenditures among the three categories or because they did not track information in that way.
• M ilitary personnel salaries. In prior study years, the study team heard from several reporting units that military R&D personnel salary data can be difficult to report as it is often tracked separately from R&D expenditures. As such, several reporting units that likely do have military personnel were unable to report salary and benefit expenditure data because they are unavailable.

Data revisions . Prior year annual data are available for FY 2022 and FY 2024. FY 2023 was not collected in an effort to improve the timing of the data collection. Moving forward, there are no plans to skip fiscal years.

In some cases, the review for consistency between the FY 2024 and FY 2025 submissions revealed significant differences due to facilities' reappraisal of their classification of various aspects of R&D programs for FY 2025. In those instances, NCSES requested that reporting units revise FY 2024 data to maintain consistency and comparability.

In cases where a response was not received in FY 2024 but was received in FY 2025, no revisions were made to the imputed FY 2024 totals. Caution should be used when comparing reported data in FY 2025 with the corresponding FY 2024 imputed amounts.

For trend comparisons, use the historical data from only the most recent publication, which incorporates changes made in prior year data to reflect corrections. Do not use data published earlier.

Changes in survey coverage and population. For FY 2025, changes to the reporting units for three agencies are described below:

• Department of the Navy. An agency-level respondent informed the survey team that several Navy research units and commands would be consolidated under one reporting unit response, an existing facility called the Naval Medical Research Command. This reduced the number of reporting units for the Navy from 18 to 12. The following reporting units were designated as inactive for the FY 2025 survey cycle:
• • Naval Health Research Center
  • Naval Submarine Medical Research Laboratory
  • Naval Medical Research Unit - SOUTH
  • Naval Medical Research Unit - INDOPACIFIC
  • Naval Medical Research Unit - San Antonio
  • Naval Medical Research Unit - Dayton

• Animal and Plant Health Inspection Service (APHIS). An agency-level respondent informed the survey team that four APHIS facilities would report under a single new reporting unit named Plant Protection and Quarantine. This reduced the number of reporting units for APHIS from seven to three. The following reporting units were designated as inactive for the FY 2025 survey cycle:
• • Forest Pest Methods Laboratory
  • Insect Management and Molecular Diagnostics Laboratory
  • Plant Pathogen Confirmatory Diagnostics Laboratory
  • Treatment and Inspection Methods Laboratory

• Food and Drug Administration (FDA). An agency-level respondent informed the survey team of several changes to FDA reporting units. The changes included renaming and
  re-organization of the following reporting units:
• • The Center for Food Safety and Applied Nutrition was renamed to the Human Foods Programs in FY 2025.
  • The Office of Regulatory Affairs was renamed to the Office of Inspections and Investigations in FY 2025.
  • The Office of the Chief Scientist is a new facility in FY 2025. Year-over-year comparisons should be made between this facility and the Office of Regulatory Affairs.

Changes in questionnaire. Several updates were made to the survey in FY 2025:

• Question 3 (types of costs) was reformatted and expanded to ask what types of expenses each facility has and whether the expense are included in their Question 2 response.
• A new question (Question 4a) was added to determine if any FY 2025 R&D expenditures came from nonfederal sources (e.g., state or local governments, businesses, higher education institutions, nonprofit organizations, or foreign sources).
• FY 2024 Question 5 (source of funds) was renumbered so that it became FY 2025 Question 4b. FY 2024 Question 4 (type of R&D) became FY 2025 Question 5.
• Instructions and definitions were clarified in questions 1, 2, 4b, 9, and 10. Reminders to cross check totals were added to questions 2, 4b, 5, 6, 8, and 10.

Facility. A unit within an agency that is responsible for performing R&D, generally with its own distinct budget and leadership. This may be a division, branch, center, laboratory, or other entity. The staff who work within the facility, and the facility itself, may be located in more than one physical location.

Fields of R&D. A list of the 56 fields of R&D reported on can be found on the survey questionnaire. In the data tables, the fields are grouped into 10 major areas: agricultural sciences and natural resources and conservation; biological, biomedical, and health sciences; computer and information sciences; geosciences, atmospheric sciences, and ocean sciences; mathematics and statistics; physical sciences; psychology; social sciences; engineering; and other fields.

Fiscal year. The federal government's financial year: FY 2025 began 1 October 2024 and ended 30 September 2025.

Full-time equivalents (FTEs). Calculated as the total working effort spent on research during a specific period divided by the total effort representing a full-time schedule within the same period. FTE R&D personnel are federal employees and military personnel only.

Public-private partnerships. Those in which the government and private companies share R&D costs.

Research and experimental development (R&D). R&D is creative and systematic work undertaken in order to increase the stock of knowledge-including knowledge of humankind, culture, and society-and to devise new applications of available knowledge. R&D has five major features:

• Novel. Advances current knowledge or creates new knowledge.
• Creative. Focuses on original concepts and hypotheses.
• Uncertain. Outcomes are not completely determined at the outset of a project.
• Systematic. Projects are planned and budgeted.
• Transferable or r eproducible. Methodology and results are transferable to or reproducible in other situations and locations.

R&D also has three major types: basic research, applied research, and experimental development.

Research, d evelopment, t esting, and e valuation (RDT&E) . Exclusive to DOD, this consists of all the activities described above for R&D conduct plus:

• Operational system development ( B udget A ctivity 7). Pre-production development of nonexperimental work on a product or system before it goes into full production, including activities such as tooling and development of production facilities.
• Software and digital technology pilot programs (Budget Activity 8). Software, electronic tools, systems, applications, resources, acquisition of services, business process re-engineering activities, functional requirements development, technical evaluations, and other activities in direct support of acquiring, developing, deploying, sustaining, enhancing, and modernizing Software Digital Technology Pilot Programs.

R&D expenditures. Money the facility spent in FY 2025 for R&D projects both within and outside a facility. Expenditures are often referred to as outlays. Expenditures include labor costs for R&D projects; noncapital purchases of materials, supplies, equipment, and services to support R&D performance; general administration costs in support of R&D activities; and expenditures for funding provided to others to conduct R&D outside a facility.

• Total FY 2025 expenditures for R&D both within and outside a facility are collected in Question 1. This includes all of a facility's R&D expenditures, regardless of whether the funding was from direct appropriations, interagency agreements, reimbursable funds, or nonfederal sources; expenditures for R&D performed by the facility's employees; and expenditures for R&D performed within and outside the facility by contractors or other entities.
• Question 2 includes total FY 2025 expenditures for R&D separated by types of costs: (a) R&D expenditures for onsite contractors, (b) all other costs for R&D performed within the facility, (c) total R&D expenditures within the facility, (d) R&D expenditures for work performed outside the facility, and (e) total R&D expenditures.
• Question 4b and all questions that follow focus on total R&D expenditures within a facility reported in Question 2, row c.

Sources of R&D funding and t ype of o rganization .

• U.S. federal government. Any agency of the U.S. government. Federal funds that were passed through from another institution were included.
• State and local government. Any state, county, municipality, or other local government entity in the United States, not including funds from state and local universities and colleges or agricultural experiment stations.
• Business. Domestic or foreign for-profit businesses or industrial firms.
• Nonprofit organizations. Domestic or foreign nonprofit foundations and organizations, except universities and colleges.
• Higher education. Domestic higher education institutions, military service academies, and consortia.
• All other organizations. Sources not reported in other categories, such as funds from foreign governments, foreign or U.S. universities, or user fees.

Type of R&D .

• Basic research. Experimental or theoretical work undertaken primarily to acquire new knowledge of the underlying foundations of phenomena and observable facts, without any particular application or use in view.
• Applied research. Original investigation undertaken in order to acquire new knowledge. It is directed primarily toward a specific, practical aim or objective.
• Experimental development. Systematic work, drawing on knowledge gained from research and practical experience and producing additional knowledge, which is directed to producing new products or processes or to improving existing products or processes. Like research, experimental development will result in gaining additional knowledge.

• Experimental development includes the following:

  • The production of materials, devices, and systems or methods, including the design, construction, and testing of experimental prototypes.
  • Technology demonstrations, in cases where a system or component is being demonstrated at scale for the first time, and additional refinements to the design (feedback R&D) following the demonstration. However, not all activities that are identified as "technology demonstrations" are R&D.

  Experimental development does not include the following:

  • User demonstrations where the cost and benefits of a system are being validated for a specific use case. This includes low-rate initial production activities.
  • Preproduction development, which is defined as nonexperimental work on a product or system before it goes into full production, including activities such as tooling and development of production facilities. Activities and programs of this type should generally be reported as investments in other major equipment.

R&D personnel. All employees who work on R&D or provide direct support to R&D, such as researchers, R&D managers, technicians, support staff, and others assigned to R&D groups or projects within the facility. Personnel may include federal employees, military personnel (civilian and enlisted), contractors, consultants, or volunteers.

• R&D personnel include all R&D personnel, whether full time or part time, temporary or permanent. Employees may perform scientific and technical work for an R&D project (e.g., designing experiments, building prototypes), plan and manage R&D projects, or provide direct support for administration of the financial and personnel aspects of R&D.
• R&D personnel does not include employees who provide general support services to the facility, such as services provided by personnel in central finance, computing, printing, maintenance, security, or similar departments in your agency that provide services to both R&D and non-R&D projects.

Types of R&D personnel .

• Researchers. Professionals engaged in the conception or creation of new knowledge, products, processes, methods, and systems and also in the management of the projects concerned. Includes R&D managers in this category.
• R&D technicians. Persons whose main tasks require technical knowledge and experience in one or more fields of science or engineering but who contribute to R&D by performing technical tasks such as computer programming, data analysis, ensuring accurate testing, operating lab equipment, and preparing and processing samples under the supervision of researchers.
• R&D support staff. Not directly involved with the conduct of a research project but support the researchers and technicians. These employees might include clerical staff, financial and personnel administrators, report writers, patent agents, safety trainers, equipment specialists, and other related employees.