COST ACCOUNTING REPORTS AND VOUCHER-BASED ACCESS TO IMPLICATIONS FOR CRS AND THE CONGRESS S. 1824 proposes that each congressional instrumentality, including CRS, prepare annual reports "detailing the cost to the instrumentality of providing support to each committee of the Senate and Senator." It would also require studies of the "feasibility of establishing a voucher allocation system for committees using the services of instrumentalities." The basic objective of these proposals is to ensure "that all resources are allocated in the most fair and efficient manner possible." All public institutions, of course, should be accountable for achieving fairness, efficiency and effectiveness in the use of resources. CRS has established many practices and continues to explore new ideas in pursuit of those objectives. It is not apparent, however, that the cost accounting and voucher proposals presented in S. 1824 would promote those objectives. This document explores implications for CRS and the Congress of cost accounting reports from and voucher-based access to CRS. First, the nature of CRS-congressional relations under the proposed system is highlighted. The next section identifies basic characteristics of CRS interactions with the Congress that would make it difficult for CRS to obtain meaningful data from the proposed cost accounting system. Additional sections discuss current and proposed disclosure of information about CRS workload in the context of the proposal to publish information obtained from the proposed cost accounting system. Another section outlines current CRS practices for fostering fairness, efficiency and effectiveness in congressional use of its resources. A concluding section identifies implications of establishing a voucher system for allocating access to congressional instrumentalities. CRS-CONGRESSIONAL RELATIONS UNDER PROPOSED COST ACCOUNTING AND VOUCHER SYSTEMS CRS currently conducts a large volume of on-going, interactive support for the Congress on a full range of public policy issues. Most transactions are completed within a day and about 90 percent within a week. Daily, interactive CRS support for the Congress would likely be a casualty of instituting the proposed cost accounting system or a voucherbased access system. The proposals would effectively transform CRS from a support agency to a contractbased consulting operation. Currently, a steady flow of support services is available through a retainer-like arrangement and is delivered on a need basis. Under the proposed systems, staff from both CRS and congressional offices would negotiate, monitor and verify costs attributed to each client for each transaction or the quantity and quality of services rendered against redemption of each voucher. The transformation of relationships between CRS and the Congress, away from interactive, on-demand services toward fully negotiated. or prepackaged units of service, would be the greatest cost of adopting these proposals. 2 CALCULATING CRS COSTS OF SERVING EACH SENATOR AND COMMITTEE In support of their cost accounting proposal, the Senate Members of the Joint Committee observe that congressional support agencies "currently track the thousands of individual requests from Members and committees through accounting and information systems for internal purposes." Thus one is left with the impression that basic data systems are in place for calculating costs of serving each congressional office. This is not the case. Moreover, the validity of data needed for making such calculations would be questionable because of production and service characteristics of CRS. The basic elements of the cost accounting process anticipated in S. 1824 are easy to understand and seem easy to implement. The application of cost accounting is easiest and most appropriate in situations in which a specific set of professional staff work on an exclusive assignment for a specified period of time to complete a relatively unique service for a single client. In such situations direct costs of completing a service can be readily determined and an allocation scheme can be designed for incorporating overhead costs. In CRS, several factors seriously complicate the application of cost accounting. These include the prevalence of joint costs in preparing and delivering products and services, highly variable and unpredictable demand for specific products and services, and high costs (and probable impossibility) of recording reasonably accurate and useful cost data. The prevalence of joint costs The prevalence of joint costs would be the most critical complication in a CRS cost accounting system. Joint costs exist when the same activity contributes to more than one product or service. The major problem arising from joint costs is determining how to make a meaningful allocation of costs across several products or services. Joint costs occur, beneficially, throughout CRS on an on-going basis because each staff member typically works concurrently on several closely related projects for several clients. This situation results from assigning work according to subject expertise. On any given work day, telephone consultations, a quick memorandum, updating an issue brief, work toward a seminar or a report may all relate, for the same subject expert, to some major public policy concern such as health care cost containment or relationships between trade and environmental degradation. Joint costs arise in each of these subject areas because work on one project typically furthers work on another. The expert is better informed for conducting a telephone consultation because of work in progress on a report. Likewise, information obtained from a telephone consultation may enhance the quality of a report in progress. Also, reading professional and trade literature and discussions with other experts often contribute to work on more than one project. Another example of joint costs is found in the effort expended with the first request received about a new development such as the resignation of a world leader, the announcement of a new Administration initiative, or a business decision with major implications for jobs, competition, or technological advancement. An initial response to the 3 first request may take hours of work. Responses to subsequent requests and preparation of any subsequent written product handled by the same subjet expert would benefit from work on earlier requests, including interactions with congressional offices in responding to those requests. The prevalence of joint costs in daily research and reference activities in CRS would make it difficult and time-consuming to allocate costs to specific products and services. Furthermore, such allocations would necessarily be based largely on arbitrary decisions. Thus CRS efforts to allocate costs across individual transactions and to aggregate costs of individual transactions by congressional client would be vulnerable to and indefensible against charges of unfairness. Variable demands across individual products and services Another complication in applying cost accounting in CRS is difficulty in forecasting demand for products or services. Reliable information for each product and service on the quantity that is likely to be delivered is needed to determine the unit costs of products and services delivered to each congressional client. There is wide variation in the number of copies of different CRS reports and issue briefs delivered to Congress, in attendance at various CRS seminars and workshops and in the use of other specific CRS products and services. Among the explanations for this occurrence is that CRS supports Congress on a full range of policy issues at all stages of the legislative process and supports representational activities of the Congress as well. Thus, a report or seminar on an issue for which policy proposals are under early stages of development might attract relatively few readers or participants. In contrast, products or services on an issue which has become a national debate topic generally enjoy a very large audience. Forecasts of how large or how small, however, are unreliable. Also, using data on the number of products and services actually delivered is complicated by the fact that the life of products or services often transcends reporting periods, in part because CRS products are frequently maintained for extended periods through revisions and updating. High record keeping costs The volume of and interactive nature of CRS support for the Congress would result in sizeable record keeping costs for CRS if it instituted a cost accounting system. In its final report, the Senate Members of the Joint Committee on the Organization of the Congress state that congressional "instrumentalities currently track the thousands of individual requests from Members and committees through accounting and information systems for internal purposes." Actually, in the case of CRS, the numbers of individual requests tracked are not thousands but hundreds of thousands each year (See Table 1). Requests are tracked only in information systems, not in accounting systems. A cost accounting system would necessary center on recording staff time, the principal element in the cost of providing CRS products and services. A staff of hundreds completes thousands of requests and services each day. Staff members may require only minutes (ignoring on-going study and preparation time of subject experts) to complete some transactions such as many telephone consultations. At the other extreme, most are also engaged in several projects on which they work intermittently over many days as time allows or as information becomes available. Given the difficulties of allocating joint costs, the high daily volume of activity, and the interspersing of work on different projects for different clients, CRS staff would have to incur substantial record-keeping costs to report reasonably accurate cost data. The effort by staff to report costs incurred for individual transactions would diminish time available for responding to congressional requests and would inescapably result in arbitrary decisions about allocating substantial costs incurred jointly in support of many projects for different clients. Implementing and mainting the proposed cost accounting system would be costly and would not help in accomplishing objectives of fairness and efficiency. COST ESTIMATES AS ALTERNATIVES TO COST ACCOUNTING: Estimating costs through one of several approaches could be pursued instead of trying to establish actual costs through cost accounting. One approach would be to develop refinements to a rather crude method, dividing the CRS budget by the number of transactions per year. Given the diversity of CRS products and services, however, weights would need to be assigned to major types, such as preparing reports, presenting seminars, providing briefings, sending prepared materials, etc. Two factors significantly complicate assigning weights. One is the prevalence of joint costs across different types of products and services. The other is great variability within each category with respect to difficulty and intensity of work required. For example, two reports of similar length often require vastly different expenditures of resources because of different research requirements based on the relative availability of information, the complexity of the topic, and the extent to which researchers have had previous experience with the topic. Another approach would be to conduct surveys of all costs incurred in sample time periods or to determine costs incurred in preparing a sampling of products and services. Again, difficulties would arise from prevalence of joint costs in preparing products and services and great variability in resources required for preparing products and services within any category. Because of these difficulties, cost estimates based on averaging or sampling processes could greatly misstate costs actually incurred in supporting individual congressional clients. For example, two clients placing the same number of requests for each of several different types of CRS products and services would appear to require the same costs in support of their needs, even if one placed relatively easy demands on CRS resources and the other always required the expenditure of substantial resources to respond to complex requests requiring especially challenging research. DISCLOSING CRS WORKLOAD INFORMATION S. 1824 would require that costs incurred in providing support to each Senator and Senate committee be published in the Secretary of the Senate's semiannual report. Provision for disclosure may be proposed because of concerns that current reporting is 5 inadequate: "...the overall demands, as well as the specific distribution of workload among the instrumentalities and the needs of individual Members and committees, are not known to the Congress as a whole." Information about the nature and volume of congressional demands placed on CRS is readily available in annual reports submitted to the Joint Committee on the Library pursuant to Section 321 of Public Law 91-510. Data from the most recent report, for fiscal year 1993, depicting CRS workload are reproduced in Table 1. The annual report also lists, by individual title and by broad subject area, more than 1,700 general distribution products prepared and maintained during the fiscal year. Not included in the CRS annual report are data depicting "needs of individual Members and committees." Disclosing information about the number and type of requests individual Members or committees place with CRS would represent a sharp departure from the CRS practice of affording confidentiality to the Congress. HOW WOULD DISCLOSING OF COST INFORMATION ASSIST IN ACHIEVING EFFICIENCY AND FAIRNESS? Efficiency and fairness in allocating the resources of congressional instrumentalities is clearly a major objective of the proposal to report costs of providing support to each Senator and Senate committee. Even if these costs could be accurately and meaningfully determined, however, it is not clear how disclosure of cost data would improve fairness and efficiency in allocating resources. To make some assessment about fairness and efficiency in the use of resources, criteria would be needed for evaluating the cost data. Among the first steps would be agreeing on the size of the Senate's share of total support costs. Within the Senate, criteria could be adopted to establish one or a combination of many alternative cost outcomes as "fair," including the following: equal costs for each Senator, committee and, possibly, subcommittee; higher costs for Senators from states with greater economic, geographic and higher costs for committees considering comprehensive legislation with broad national interests (an omnibus trade, bill, health care reform proposals) or a large volume of legislation, etc. |