| Federal
Committee on Statistical
Methodology Office of Management and Budget |
FCSM
Home ^ Methodology Reports ^ |
Statistical Policy Working Paper 8 - Statistical Interagency Agreements
Click HERE for graphic. MEMBERS OF THE FEDERAL COMMITTEE ON STATISTICAL METHODOLOGY (October 1981) Maria Elena Gonzalez (Chair) Office of Information and Regulatory Affairs (OMB) Barbara A. Bailar Bureau of the Census (Commerce) Norman D. Beller National Center for Education Statistics (Education) Yvonne M. Bishop Energy information Administration (Energy) Edwin J. Coleman Bureau of Economic Analysis (Commerce) John E. Cremeans Bureau of industrial Economics (Commerce) Zahata D. Doering Defense Manpower Data Center (Defense) Marie D. Eldridge National Center for Education Statistics (Education) Daniel E. Garnick Bureau of Economic Analysis (Commerce) Charles D. Jones Bureau of the Census (Commerce) Daniel Kasprzyk Social Security Administration William E. Kibler Statistical Reporting Service (Agriculture) Thomas Plewes Bureau of Labor Statistics (Labor) Raymond C. Sansing Internal Revenue Service (Treasury) Fritz J. Scheuren Internal Revenue Service (Treasury) Monroe G. Sirken National Center for Health Statistics (RES) Wray Smith Energy Information Administration (Energy) Thomas G. Staples Social Security Administration (HHS) OFFICE OF INFORMATION AND REGULATORY AFFAIRS Christopher DeMuth, Administrator Thomas D. Hopkins, Deputy Administrator for Regulatory and Statistical Analysis Maria E. Gonzalez, Chairperson Federal Committee on Statistical Methodology PREFACE The working paper was prepared by the members of the Subcommittee on Statistical Interagency Agreements, Federal Committee on Statistical Methodology. The Subcommittee was chaired by Monroe G. Sirken, National Center for Health Statistics, Department of Health and Human Services. This report presents data on the prevalence of statistical interagency agreements. It also summarizes the prevailing practices and policies of Federal agencies with respect to these interagency agreements. The information was compiled in a pilot survey involving a sample of Federal agencies that served as sponsors and contractors of statistical interagency agreements during 1979. These subcommittee conclusions were not formally endorsed by the Federal Committee on Statistical Methodology nor by the Office of Management and Budget. The summary presented here provide guidance for the use of statistical interagency agreements. Meetings will be organized to discuss the findings of this subcommittee with Federal agency personnel involved with statistical interagency agreements. i MEMBERS OF THE SUBCOMMITTEE ON STATISTICAL INTERAGENCY AGREEMENTS Monroe G. Sirken, Chairperson National Center for Health Statistics (RES) Paul Becker Administrative Programs and Services (Labor) Ralph W. Gann Statistical Research Service (Agriculture) Maria E. Gonzalez Office of Information and Regulatory Affairs (OMB) Sue Lindgren Bureau of Justice Statistics (Justice) Dawn Nelson Bureau of the Census (Commerce) David B. Orr National Center for Education Statistics (Education) Nathaniel Pigman, Jr. Health Care Financing Administration (HES) Anitra Rustemeyer Energy Information Administration (Energy) Thomas C. Walsh Bureau of the Census (Commerce) ACKNOWLEDGMENTS The data collected on statistical interagency agreements from sponsor and contractor agencies was based on questionnaires prepared by the subcommittee, especially by Paul Becker, Ralph Gann, Sue Lindgren, Sue Rustemeyer and Tom Walsh. The cooperation of the agencies included in this study was essential to the results presented. The report was reviewed by all members. The sections were initially prepared by: 1. Introduction. . . . . . . . . . . . . . . . . . . Maria Gonzalez 2. Design of the Study . . . . . . . . . . . . . . . Maria Gonzalez 3. Interagency Agreements During 'FY 1979. . . . . . . Monroe Sirken 4. Policies and Practices Affecting Reimbursable Work. . . . . . . . . . . . . . . . . . Dawn Nelson and Tom Walsh 5. Summary and Conclusions. . . . . . . . . . . . . . . Monroe Sirken STATISTICAL INTERAGENCY AGREEMENTS* 1. INTRODUCTION A Federal agency may elect to conduct its statistical work in- house, have the work performed by another Federal agency (interagency agreement), or contract the work to a non-Federal organization; or, it may use any combination of these three mechanism. Previously, little was know about interagency agreements and currently no organized body of Federal regulations and practices exist concerning such agreements. Basically, any interagency agreement is a written agreement in which both partner--the sponsor and the contractor--are Federal agencies..1 The agreement arranges for the transfer of funds in payment for services and products such as the sharing of computer time or the conceptualization and performance of a major project. Congressional authority for entering into interagency agreements is authorized by the Economy Act of 1932, as amended (31 U.S.C.686). That Act allows any Federal agency to contract with another Federal agency to contract with and recover all or part of the cost incurred of the contractor agencies surveyed contains allowance for providing information to customers upon request (e.g., Commerce, HHS, Labor). The Subcommittee on Interagency Agreements of the Federal Committee on Statistical Methodology (FCSM) recently completed an investigation of the feasibility of conduction a survey of Federal agencies to collect information on statistical interagency agreements. This article summarizes the work of the Subcommittee. For purpose of the study the Subcommittee defined statistical work as any of the following activities: sample design or selection; survey methodology testing and research; data collection; data processing; sample weighting; tabulations; variance estimation; data analyst; statistical consultation and training; and publication of results. An example of work carried out under a statistical interagency agreement is the National Health Interview Survey (NHIS), a survey designed to measure the health conditions, practices and services of the population. Is sponsored by the National Center for Health Statistics through a statistical interagency agreement with the Bureau of the Census. The Subcommittee deliberated on the needs for information about statistical interagency agreements. Two kinds of data requirements were identified: (1) statistic on the prevalence and characteristic os statistical interagency agreement. The feasibility of compiling this information was tested by conducting surveys of Federal agencies. The results of the feasibility study collected information was used to estimate the total number and the dollar value of such projects and map merits. The information on interagency agreements and to make suggestions for future study in this area. The balance of his article provides information on the design of the feasibility study; the dollar value, number and network of agencies involved in interagency agreements; an overview of the responses to questions on administrative agency agreements. The report concludes by summarizing what has been learned from this project and the implications for the future. 2. DESIGN OF THE STUDY The feasibility study of statistical interagency agreements covered both contractor agencies and sponsoring agencies. Contractor agencies carry out the statistical work specified and paid for by the sponsoring agencies. These agencies are referred to as the contractor and the sponsor, respectively. It should be noted that an agency can be both a sponsor of some interagency agreement and a contractor for others. For example, the Bureau of Labor Statistics sponsors the CPS which is carried out buy the Census Bureau and is the contractor for an International Development to train foreign nationals in labor statistics. The study collected data from nine contractors and nine sponsors. Although the sponsor and contractor questionnaires were distinct, they were designed to be comparable on several points. Both questionnaires included questions on policies and practices for interagency agreements (either sponsored or contracted, as appropriate) including selected information regarding the agreements..2 Questions on criteria for entering into interagency agreements covered such topics as confidentiality, cost, quality, _____________________________________ *The Subcommittee on Statistical Interagency Agreements of the Federal Committee on Statistical Methodology was chaired by Monroe G. Sirken, National Center for Health Statistics. The members of the subcommittee include Paul Becker (Labor), Ralph Gann (SRS, Agriculture, Maria Gonzalez (OMB), Sue Lindgren (BJS/Justice), Dawn Nelson (Bureau of the Census, Commerce), David Orr (NCES/Education), Nathaniel Pigman, Jr. (HCFA/HHS) and Thomas C. Walsh (Bureau of the Census, Commerce). .1 Federal departments are the executive administrative divisions of government such as the Department of Commerce or the Department of Agriculture. We will designate as Federal agencies any subdivisions of a department such as the Bureau of the Census or any independent agency of the Executive Branch of Government. .2 Specimen copies of the contractor and sponsor questionnaires are available upon request. 1 timeliness, and types of funds available. Questions on the process for implementing interagency agreements covered technical specifications, personnel ceilings and OMB requirements for reimbursable surveys. The agencies selected for the feasibility study, as well as their departmental affiliation are listed in Exhibit 1. Exhibit 1 CONTRACTS AND SPONSORS OF INTERAGENCY AGREEMENTS IN THE FEASIBILITY STUDY (BEA) Bureau of Economic Analysis, Department of Commerce (BLS) Bureau of Labor Statistics, Department of Labor (CENSUS) Bureau of the Census, Department of Commerce (ESCS) Economics, Statistics, and Cooperatives Service, Department of Agriculture (HCFA) Health Care Financing Administration DHEW (SSA) Social Security Administration, DHEW SPONSORS (AID) Agency for International Development, Department of State (EDA) Economic Development Administration Department of Commerce (EIA) Energy Information Administration Department of Energy (HCFA) Health Care Financing Administration DHEW (HUD) Assistant Secretary for Program Development and Research Department of Housing and Human Development (BJS) Bureau of Justices Statistics, Department of Justice (NIOSH) National Institute for Occupational Safety and Health, Public Health Service, DHEW (NSF) National Science Foundation (WPRC) Water and Power Resource Center, Department of the Interior Questionnaire were mailed with cover letters addressed to the directors of the selected agencies. Whenever possible the letter suggested the name of a person to serve as liaison with the subcommittee. These person were members of the subcommittee who worked in the selected agencies. Contractor questionnaires were mailed on June 15, 1979, and responses were requested on July 20, 1979 and responses were requested by July 20, 1979. Sponsor questionnaires were mailed on January 9, 1980, and responses were requested by February 8, 1980. All agencies responded. The reference year for listing statistical interagency agreements was FY 79. The nine contractors selected for the feasibility study are responsible for a very large proportion of the statistical work done through interagency agreements..3 Agencies serving as contractors for large volumes of interagency statistical work were purposively selected. Therefore, their listings of interagency agreements provide fairly complete coverage of interagency statistical projects for FY 79. These contractor listings included the name of the agency sponsoring the agreements. From the contractor listings, the subcommittee made a judgmental selection of nine sponsoring agencies which included a variety of departments and the major sponsors in terms of the amount of statistical interagency work conducted. However, the nine sponsor agencies represent only a small portion of all agencies which sponsor interagency statistical work. Only nine agencies were include because the study was primarily focus on the feasibility of collection data on interagency agreements. Also, if less than ten cases are surveyed OMB clearances are not required. The information based on e sponsor agencies cannot be generalized; it serves to illustrate the policies and practices affecting interagency agreements. 3. INTERAGENCY AGREEMENTS DURING FY 1979 based entirely on the information reported by the nine contractor agencies that were reported for each of their 1979 interagency agreements: 1. Name of sponsor 2. Cost of interagency agreement 3. Cost of subcontracts 4. Statistical activities performed The nine contractor agencies reported a total of 179 interagency agreements involving over 25 sponsor agencies. The total dollar value of these agreements reported by the contractor agencies was almost $100 million, or roughly one-tenth of the total Federal statistical budget. Thirty-five agreements involved subcontractors with a combined dollar value of $8 million. The Bureau of the Census was by far the biggest contractor. It was the contractor on 92 out of 179 agreements. BEA, BLS, and NCHS were contractors for about 20 agreements each, and ESCS was the contrator for fewer than 10 interagency agreements each The average dollar value of interagency agreements during 1979 was about $550,000, but the average varied considerably among the contractors. Census and BLS had the highest averages, $780,000 and $650,000, respectively. The smallest averages were under $50,000 for SSA ___________________________________ .3 The chapter on Interagency Funding of A Framework for Planning U.S. Federal Statistics for the 1980's (Department of Commerce, 1978) points our agencies doing a major part of statistical interagency work and some of the work. The contracting agencies were selected taking into account these agencies. The contractors selected for this feasibility study carry out the major portion of interagency statistical agreements, probably about 90 percent or more. 2 and HCFA. The averages for the remaining agencies ranged from about $100,000 for BEA to somewhat over $300,000 for NCHS and IRS. The interagency agreement is in large part due to differences in kinds of functions carried out by the contractor. The Census Bureau, for example, often performs a full complement of survey related functions including conducting large national population and establishment surveys by personal interview Interagency agreement with the Census Bureau add to almost $72 million or about 70 percent of the total dollar value of all interagency agreement reported by the contractor agencies. Agreements with BLS and NCHS account for about 13 percent and 5 percent, respectively, of the dollar value of all agreements. Interagency agreements with BEA, IRS, and ESCS each represent about 2-3 percent. The dollar value of agreements involving HSFA and SSA as contractors was less than $100,000. The contractor agencies subcontracted some work on about one- fifth of the interagency agreements, but the dollar value of the subcontracts represented less that 10 percent of the dollar value of interagency agreement. Several contractors, including Census, IRS, and BEA did not subcontract any part of their interagency agreements 1. Sample design and/or selections 2. Methodological testing and research 3. Data collection 4. Data processing 5 Tabulations 6 Variance estimates 7. Data analyses 8. Publication of results 9 Statistical consultations The proportions of interagency agreements that involved each type of activity are distributed by the contractor agency in Table I. Data processing and tabulations were the most often reported activities. About 75 percent of the agreements involved each of these activities. Statistical consulting per se was the actively reported least often (four percent, and publication results involved about 35 percent of the agreements. There are substantial differences among the contractor agencies. For example, sample design activity was involved in most census and BLS agreements but in only 5 percent of BEA's agreements. Only three contractor agencies provided the full range of statistical activities. They are the full range of statistical activities. They are the Census Bureau, BLS, and NCHS. The number and dollar value of interagency agreements as reported by the nine contractors are distributed in Table 11 by the sponsor agencies. As sponsors, BLS and BUS spent far more through interagency agreement than other agencies. BLS spent about $20 million and BUS spent almost $10 million. The dollar value of interagency agreements was less than $5 million for each of the other sponsors agencies. It is interesting to note that note that some contractors were also large sponsors. The dollar value of interagency agreements sponsored by BLS is nearly $20 million or almost equal to the value of agreements for which is served as the contractor, and neither statistical agreements during fiscal year 1979. The dollar value of statistical interagency agreements for fiscal year 1979 are distributed by both the contractor and the sponsor agencies in Table III. The table shows the amount of money transferred between specified sponsors and contractors of interagency agreements. BEA and Census had the most widespread network of interagency agreements. The Census Bureau served as nearly all the listed sponsor agencies. BEA and agency agreements. The Census Bureau served as neraly all the listed sponsor agencies. BEA served as contractor for 14 of the 24 listed sponsors. for other agencies were far more restricted both in terms of the number and dollar values of the agreements. BLS served as contractor for interagency agreements with nine agencies, but more that 90% or the dollar volume was with other agencies in the Department of Labor.NCHS served as contractor for six agencies but virtually all the dollar value was accounted for by agreements with agencies in the Public Health Service. Each of the other contractor agencies had interagency agreements with fewer than five sponsor agencies. The statistics presented in this section should be interpreted with caution since they are the by-product of a feasibility test of the survey method. The reported statistics understate the number and the dollar volume of statistical interagency agreements, since coverage in the pilot was limited to nine contractor agencies sponsor agency. It is believed, however, that the nine contractor agencies were identified by the sponsor agencies in the pilot study. They are the Customs Bureau, U.S. Bureau of Standards, U.S. Geological Survey, and the Office of the Assistant Secretary of Planning and Evaluation, DHEW. Based on the reports of the sponsor agencies was less than $1 million in fiscal year 1979. 4. POLICIES AND PRACTICES AFFECTING INTERAGENCY WORK The contractors and sponsors in the study were asked about the factors they considered in entering into and carrying out statistical activities with another Federal agency on a reimbursable basis. This information was elicited to provide insight into the perceived advantages and disadvantages of interagency agreements in contrast with private sector contracts for statistical services. The responses to these question on policies and practices are summarized in this section. A. Contractor Agencies i. Policies for conducting interagency work: Among the contractor agencies, there was a consensus that statistical work conducted for another agency on a reimbursable basis must be consistent with the mission, major goals, and legislative authority of the contractor agency. The effects of such policies impose certain limitations on the nature of the surveys and other statistical activities in which an agency may engage, depending on the scope of the contractors mission. An agency whose mission is broadly defined has the liberty to conduct a greater variety of statistical services than one whose mission is more narrowly defined. Most contractors also mentioned that reimbursable work could be undertaken only if it did not interfere with the agency's ability to perform its regular duties. Several of the agencies performed is usually small enough to avoid any serious impact on their own goals and programs. One Census, but generally applicable to all Federal agencies, is that the Bureau will not deliberately make efforts to compete with the private sector. For example, the Bureau will not respond to another government agency's general solicitations for bids for conducting a particular statistical activity. The study also explored what limitations, if any, contractors place on the type of information they are willing to collect, particularly with regard to highly sensitive subjects. Answers to this inquire varied among the agencies indicating reared this subject. The general conclusion is that contractors will engage in the collection mission and is regarded by them to be in the public interest. A number of agencies did state sensitivity of the subjects being considered and the possible adverse effects certain subjects might have on the agency's regular programs. A few other conditions for conducting statistical activities under an interagency agreement to specify that appropriate standards of accuracy and quality be employed and maintained throughout the project. Certain sponsor regarding the publication of the results, limitations of the data. Most contractor agencies stressed the importance of requiring contractors reserve the right to publish data under certain conditions specified in the interagency agreement. 5 ii. Confidentiality Policies and practice regarding confidential treatment of information collected under interagency agreements were also examined in the study. There is no general legislation which provides uniform protection of confidential statistical data for all Federal agencies. Virtually all of the contractors, however, cited agency policy and/or law which protects the confidentiality of respondents when disseminating information to users, including the sponsor. The laws referred to basically protect the identity of an individual or establishment reporting confidential information either by limiting data access to authorized persons, or by restricting the use of the data for statistical purposes. Agencies whose statistical activities are not covered by such laws have strong written policies and the information reported by them. A statement on the agency's confidentiality policy is contractors, and it is usually noted that the data set may have to be modified for confidentiality reasons prior to its release. iii OMB clearance and personnel ceilings: Federal agencies are required to obtain approval from the Office of Management and Budget (OMB) for data collection activities which involve more than nine respondents. The sponsor is generally responsible for obtaining the approval, but it requires close coordination with the contractors to prevent delays and scheduling disruptions. Many interagency projects involve new survey applications and introduce new requirements or procedures which require extensive justification and explanation by both agencies. Even if the sponsor is experienced in this process, it generally takes longer when both agencies are involved in the preparation and approval of the clearance package before submission to OMB. The time required to obtain clearance is critical factor because it inhibits the contractor from establishing a definite worked schedule around the agency's operational requirements and can reduce the amount of time the contractor has to expend single-year funds which interagency agreements between agencies. A contractor agency may have some positions identified perform work for recurring agreements. With these personnel limitations, contractors cannot seek additional interagency agreements and, at times must decline or delay expanding work on existing agreements. Decreases in the permitted ceiling levels have forced contractors to service only those sponsors with high priority work which complements existing programs of the contractor. iv Subcontracting The contractor agencies were queried on the extent to which subcontracts with other organizations were used to obtain certain services required in completing the work under an interagency agreement. Such practices seem to vary depending on the mission of the agency. Several contractors reported that the types of interagency agreements entered into usually call for services which they are most capable of rendering; thus, there is no need to subcontract with other organizations. A few agencies reported following the practice of subcontracting with other organizations to perform certain functions such as data collection, data processing, and analysis where there is a clear cost benefit to the sponsor or when they do not possess the capabilities or resources t perform the functions within the specified time frame. v. Contractor agency advantages In summarizing their experiences with reimbursable statistical work, the contractor agencies were asked what advantages, if any were derived from engaging in such interagency agreements. Unanimously, the responses were positive, pointing out several benefits to their agencies as well as to the statistical community as a whole. Interagency work created an open flow of communication between members of the Federal statistical community, resulting in knowledge and expertise being shared efficiently. Such work allows for increased coordination within the Federal statistical system, leading to a greater awareness of the data needs among agencies; increased uniformity in survey concepts, procedures, and design; and a reduction in duplication of data collection efforts, as well as reductions in overall respondent burden. Although the contractor agencies expressed some difficulties in executing work under interagency agreements due to certain restrictions regarding funding, personnel, and other administrative matters, the overall opinion of those surveyed seemed to be that such work enhances a contractor's ability to serve the statistical community and to carry out its own designated mission. vi Funding: After this feasibility study was completed, the budget situation throughout the Federal Government changed considerably, generally resulting in less money for reimbursable statistical work. As a consequence, funding has become more of a concern for contractor agencies. Contractors are experiencing difficulties in planning their work because of the uncertainty of funding from sponsors. Much of the work undertaken through an interagency agreement, for example, survey implementation, requires considerable advance notices and adequate preparation time. Recently, contractors such as the Census Bureau have been forced to start or continue work which may eventually be sharply curtailed if the sponsor is not able to provide the expected level of funding. If the contractor decides to wait until the sponsor's funding is certain, considerable delays may result in completing the work..4 B. Sponsor Agencies i Reasons for interagency agreement: A few agencies that have sponsored interagency statistical work were queried on the factors they _________________________________ .4 For additional information on the effects of recent reductions in the statistical budgets of Federal agencies, see Daniel Melnick, et. al., Recent Changes in the Federal Government's Statistical Programs: An Overview of the President's Budget for FY 1993 and Analysis of the Departments of Energy, Labor and the Bureau of the Census, (Washington, D.C.: Congressional Research Service, The Library of Congress, April 8, 1982). 6 considered in deciding to have such work performed by another Federal agency and on their experiences with such agreements. Sponsors included in the study indicated that consideration was given, first, to the amount and type of resources available within their own organization in view of the tasks required to carry out the statistical project. The decision to contract the work is influenced by factors such as the size and expertise of staff required, accesibiltiy to a convenient sampling frame, current agency. Often, a sponsor whose primary mission is not one of a statistical nature has a need for various data, but does not carry out the necessary functions to produce such information. Thus, the need arises to contact with another organizations (Federal agency or private firm) to provide the desired statistical data. The sponsor agencies were also asked what factors influenced their decisions to have statistical work performed by another Federal agency instead of a private organization. The sponsors cited several factors which often led them to contract with another Federal agency. Among the more common reason given were the unique qualifications offered by certain Federal agencies such as the ability to conduct large scale survey because of the availability of a national sampling frame, a permanent regional field staff, and general expertise in this area. Similarly, NCHS would be a prime candidate for performing work in health-related subject matter areas, whereas SSA would be chosen for obtaining special tabulations from the various administrative record systems (Social Security, Medicare, etc.) under its control. In addition, for projects of a continuous nature, there are some clear advantages to be gained from the continuity of procedures, personnel, and experiences offered by the Federal engaged to conduct the work. Cost and possible access to other related information possessed by the contractor agency were also mentioned as reasons which influences a sponsor's decision to have the work conducted within the federal establishment. ii. Funding and OMB clearance Very often reimbursable statistical work between Federal agencies is funded with single year appropriations; i,e., funds appropriated for a given fiscal year which must be obligated during that year. There are several exceptions to this practice, however, usually depending on the type of appropriations received by the sponsor agency or the source from which funds for the statistical activity are obtained. A few sponsors reported having 2-year funding which allows them to obligate funds in the year appropriation and also during the following fiscal year. Additionally, there are a few situations in which the funds designated for a statistical activity need not be obligated in a given time frame (no-year funding); however, this appears to be rare for statistical activities. OMB clearance creates timing problems for the sponsor agency as well as for the contractor. Often the sponsor is unfamiliar with the OMB process, including certain Privacy Act requirements, and may not be prepared to provide the detail required or to spend the time needed to obtain clearance. When more time is required than expected to obtain clearance, work schedules are delayed and funds may be lost if they cannot be obligated within the time allowed. Work can also be delayed or canceled if the sponsor has failed to include a data collection project in the agency's Information Collection Budget (ICB). The ICB describes each existing and proposed new information collection effort to be implemented or continued during the next fiscal year and must be approved by OMB. This provides a method of controlling the response burden created by a project before OMB clearance can be obtained. iii. Confidentiality: Several of the sponsors reported that certain types of analyses, in particular microdata studies, and other forms of statistical activity requiring the user's access to microdata files, were hampered or prevented by the constraints of the contractors confidentiality laws and practices. The activities mainly affected by such restrictions are small area data studies matching to other data studies, matching to other data sets, and use of survey records for follow-up studies. Some of this problems have been overcome by having the contractor perform the desired activities such as producing small area data tabulations; however, this usually results in higher costs to the sponsor and sometimes requires more time than are not so easily resolved due to different practices and perceptions among agencies regarding privacy protection, disclosure risks, iv. Cost, timeliness, and quality: Sponsors reported mixed results in experiences with statistical work performed by other Federal agencies. With regard to costs, most sponsors were of the opinion that costs of the work conducted by Federal agencies measured quite favorably, possibly even lower than costs charged by private contractors, although direct comparisons were not possible in most instances. One noteworthy point was that often the desired data could be obtained from another Federal agency in conjunction with the collection efforts of an existing survey, thereby substantially reducing costs to the sponsor. In addition efficient sampling frames often are available only from another Federal agency responsible for administrative record systems pertaining to certain specific populations. When questioned on the timeliness of work performed by other Federal agencies, virtually all the sponsors in the study reported that contractors in the Federal sector usually took longer to complete a job than did their counterparts in the private sector. One explanation offered for this observation was that priority to their own program requirements which could bear on the timing of completing work for other agencies. The quality of the work performed by Federal agencies compared favorably with that or private contractors, and several sponsors indicated that 7 they felt it was better. Again, direct comparison of the work conducted by the groups were not possible; thus, these comments reflect only the impression of sponsor agencies included in this study. It as the consensus of these sponsors that the statistical work performed by Federal contractors was of a high quality and that most of them engage in the type of statistical activities which suit their particular expertise and experience. 5. SUMMARY AND CONCLUSIONS An agreement between Federal agencies that involved the transfer of funds from a sponsor agency to a contractor agency to perform the specified work is called an interagency agreement. Although interagency agreements account for a substantial part of the Federal statistical workload, little is known about how widespread this arrangement is, or what the prevailing policies and practices are. Little attention has been given to how this information might be gathered. This report presents preliminary findings from a small pilot study that was serve as contractors and nine statistical agencies that serve as sponsors of interagency agreements. The pilot survey findings on the prevalence of statistical agreements during 1979 are summarized below; (1) Volume of interagency work There were about 180 interagency statistical agreements during 1979. They involved about a dozen contractor agencies and several times that number of sponsor agencies. The dollar volume of interagency finding was about $100 million (2) Work distribution among contractor agencies The Bureau of the Census accounted for over 70 percent of the dollar value of reported interagency during 1979. Virtually all of the other reported interagency work was performed by five statistical agencies: National Center for Health Statistics alone sponsored about 22 percent of the dollar volume of interagency agreements. Other large sponsors of interagency work were the Department of Health, Education and Welfare (18 percent) and the Department of Justice (12 percent). (4) Contractor-sponsor networks The Bureau of the Census had the widest network of interagency statistical agreements. It performed interagency work for sponsor agencies in virtually every department of government. The Bureau of Economic Analysis also had agreements with agencies in many departments. The networks of other contractors were not as extensive and tended to be limited that were located in their own departments. Prevailing policies and practices reported by contractors and sponsors in the survey ar summarize below. (1) Conditions favoring interagency agreements Contractors agencies give first priority to interagencies projects with objectives most closely related to their own. They do not undertake work that is incompatible with their own mission. When sponsor agencies lack personnel or logistical resource but have the funds to finance statistical work they turn to interagency agreements or to contracts with non-government agencies. Interagency agreements are especially attractive if the contractor agency has uniquely suited qualifications or resource such as surveys, to link the collection of data to an existing data base, or to access the appropriate sampling frames (2) Benefits of interagency agreements Interagency work facilitates the coordination of statistical programs within the Federal statistical system. It leads to more awareness of common data needs among agencies, and to greater uniformity in concepts, definitions and procedures. It leads to more awareness of common data needs among agencies, and to greater uniformity in concepts, definitions and procedures. It leads to overall reductions in respondent burden when, for example, the sponsor's survey is piggy-backed onto the contractor's existing survey. It often enhances the statistical programs of contractor as well as sponsor agencies. (3) Restrictions on interagency agreements Federal regulations favor a policy for utilizing the private sector, insofar as possible, to provide the products and services necessary to support government functions. In compliances with this policy, Federal agencies avoid competing with the private sector contracts include: funding for interagency agreements is often limited to a single fiscal year, government imposed personnel ceiling limit the resources available to agencies, and sponsor seriously hampered by the confidentiality constraints of contractor agencies. (4) Comparisons with private contract The costs and quality of statistical work performed through interagency agreements appear to compare favorably with the work performed by the private sector. Response rates, for example, are generally higher in surveys conducted by the Bureau of the Census than in those conducted by non-government agencies. Furthermore, interagency agreements avoid delays normally associated with the competitive contract process. On the other hand, the private sector is often perceived by the sponsor agency as more dedicated than Federal contractors to the sponsor's needs and interests. So much for the summary of the substantive findings. Turning attention now to the methodological findings, the pilot study experience indicated that the agency survey was better suited to collect statistics on the prevalence of 8 interagency agreements than to compile information on prevailing policies and practices. The agency survey is not particularly well suited for collecting information about polices and practices of interagency agreements. These policies and practices involve complex and even controversial issues that did not readily avail themselves to the instruments and methods that were used in the pilot study. Another approach would be to arrange interagency meetings of sponsor and contractor agencies at which the joint staffs can informally discuss, review and evaluate current practices and policies and, as needed, revise them for the common good of the Federal statistical system. On the other hand, it is quite feasible to compile statistics on interagency agreements by conducting surveys with a relatively small number of the largest statistical agencies since they serve as contractors for a very large fraction of all such interagency agreements. It would be of the largest agencies with a relatively small sample of other contractor and sponsor agencies in order to check the completeness and the quality of the information reported by the largest agencies. Also, further developmental work is indicated to improve the survey instruments. In particular, definitions and concepts such as interagency agreement, statistical project, etc., need to be sharpened the most efficient way of compiling statistics about interagency agreements, especially if their funding levels charge substantially from year to year which appears likely to be the case. For agreements presented in this report are probably in the overall Federal statistical budget. If needed on a fairly regular basis, serious considerations should be given to compiling them as a by-product of an existing information systems such as the OMB Reports Management System or the Federal budget process. It is the sense of this Committee that programmatic statistics, such as those presented in this report would be invaluable in planning and coordinating the Federal statistical system. Reports Available in the Statistical Policy Working Paper Series 1. Report on Statistics for Allocation of Funds GPO Stock Number 003-005-00178-6, price $2.40 2. Report on Statistical Disclosure and Disclosure-Avoidance Techniques GPO Stock Number 003-005-OO177-8, price $2.50 3. An Error Profile: Employment as Measured by the Current Population Survey GPO Stock Number 003-005-00182-4, price $2.75 4. Glossary of Nonsampling Error Terms: An Illustration of a Semantic Problem in Statistics (A limited number of copies are available from OMB.) 5. Report on Exact and Statistical Matching Techniques GPO Stock Number 003-005-00186-7 price $3.50 6. Report on Statistical Uses of Administrative Records GPO Stock Number 003-005-00185-9, price $5.00 7. An Interagency Review of Time-Series Revision Policies (A limited number of copies are available from OMB.) 8. Statistical Interagency Agreements (A limited number of copies are available from OMB.) Copies of these working papers, as indicated, may be ordered from the Superintendent of Documents, U.S. Government Printing Office, Washington, D. C. 20402.
| Page Last Modified: April 20, 2007 | FCSM
Home Methodology Reports |