Click for next page ( 20

The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement

Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 19
4 Suggestions for Improving the Early Estimating Procedures for Federal Agencies ,. As discussed in the preceding chapters, inaccu- ste estimates are not the only cause of budget- related problems on federal projects, and the proce- dures currently used by federal agencies to prepare early estimates are not inferior to the procedures used by most private owners. Nevertheless, the committee age that federal procedures for pre- paring early estimates could be and should be im- proved, and the committee has prepared a list of suggestions for the consideration of the agencies. These suggestions, which are discussed below, were distilled from a large number of ideas generated in the course of the committee's discussion. Some of the suggestions tend to complement one another, while others are unrelated. INTERAGENCY COOPERATION Estimating is a field in which a free and continu- ous flow of information is vital- as illustrated by the flow diagram in Figure 1, Chapter 2. Federal agencies are in a unique position to interchange information to their mutual benefit; however, indi- cations are it is an underused opportunity. The committee believes that there are four areas in par- ticular in which agencies could benefit from greater cooperative action: standardization of tenninol- ogy, standardization of estimate formats, joint re- search efforts, and pooling of cost data. Standardization of Terminology As shown in Appendix E and as discussed in the preceding chapter, different federal agencies use different terminology for cost estimates. While the differences in terminology have not caused any serious problem to the committee's knowledge, they certainly have made the committee's task much more difficult and have probably created problems for others who work with budgets and estimates for a number of different agencies (e.g., members of Congress, congressional staff personnel, Office of Management and Budget personnel, and private A- E firms). The absence of consistent terminology probably has also contributed to the general lack of interagency cooperation that the committee ob- served, particularly with regard to the collection and use of cost data. The committee suggests-that agencies agree on terminology for construction cost estimates and budgets. The agencies might con- sider, for example; adapting the terms that the committee has used, which are defined in Chapter 1. In any event, conflicting use of the same termi- nology should be eliminated where possible. Standardization of Estimating Formats There has been considerable discussion among estimators for a number of years about the relative merits of trade (CSI) and elemental (UNIFORMS methods of organizing construction estimates. Proponents of the elemental method believe it is the best method for organizing estimates because costs can be presented in terms of building systems, which is in accord with the way most engineers, architects, and owners think of a building. Propo- 19

OCR for page 19
20 nents of the ;rade approach note that most con~ac- tors use Me CSI format, and furthermore, constnuc- tion specifications are assembled using a CSI se- quence. The committee believes that both ap- proaches have merit, but the committee suggests the use of the elemental method for early estimates and the trade method for the later and more detailed estimates. Automated estimating systems like CACES permit estimates prepared in one format to be translated into the other format. Joint Research Efforts The committee has seen relatively little evidence of interagency cooperation on research related to cost engineering or the sharing of the results of research carried out by individual agencies. Agen- cies discuss their research work in Federal Con- s~uction Council meetings, but they do not initiate any joint research activities Trough the council or make extensive use of information that is presented at council meetings. For example,. the.committee understands mat no agency has adopted a computer estimating system developed by another agency (e.g., the 1391 Processor or CAGES). The committee saw a great deal of interest on the part of the agencies in cost engineering in gen- eral and a desire to improve the accuracy of their cost estimates. However, agencies seem to have differing views on how estimates should be pre- pared and the need for or value of research. Agen- cies also probably suffer from the "not invented here" syndrome. The committee believes Hat agencies would benefit greatly if they could overcome impediments to cooperative research efforts and the sharing of research results. Joint sponsorship of research might help eliminate the `4not invented here" problem by giving all the participating agencies a feeling of ownership of research results. Pooling of Cost Data The committee also has observed Hat although historical cost data are kept by most agencies, such information is seldom shared by the agencies ex- cept in the Department of Defense. Some agencies have expressed the view Hat the pooling of cost data would be of little value because different agen- cies construct very different facilities. There may be some truth in this argument. For example, the health care facilities construct by He Indian Health Service are very different from the health care fa- EARLY COST ESTIMATES FOR FEDERAL CONSTRUCTION PROJECl S cilities constructed by the Veterans Administration and the Department of Defense. On the other hand, the large hospitals constructed by the Veterans Administration would appear to be generally com- parable to the large hospitals constructed by the Department of Defense, and the cost data of one agency should be usable by the other agency. Similarly, although the industrial plants of the Deparanent of Energy are unique to that agency, other facilities constructed by DoE (e.g., laborato- ries, office buildings, and warehouses) are not unique, and DoE data on the cost of such facilities might be of benefit to other agencies. Finally, at the elemental level He cost of an item is unrelated to the type of facility being constructed; thus, data on the cost of brick, for example, could easily be shred. While it is possible that the idea of pooling cost data on facilities constructed by many different agencies might prove impractical for administra- tive or organizational reasons, the committee be- lieves that the idea should at least be explored. THE IMPORTANCE OF A-E ESTIMATORS In recent years federal agencies have come to rely on the private sector for most activities related to design of federal facilities, including He prepa- ration of estimates. Thus, a large percentage of the early estimates used by federal agencies to request funds from Congress are prepared by estimators employed by A-E firms or by independent consult- ants who are subcontractors to A-E fogs with prime design contracts. These private estimators also develop most of the subsequent higher level esti- mates that are prepared to help ensure that a project stays within budget. Since, as discussed previously, accurate estimates are one of the keys to a successful project, the qualifications of estimators employed by A-E firms is a legitimate concern of the agencies. However, to the committee's knowledge the qualifications of the estimators to be used by an A-E are seldom a major consideration in the selection of an A-E firm for a design contact. In fact, it is sometimes not even discussed. The committee feels that this is a serious shortcoming of the A-E selection process and that cost estimating capability should be more heavily weighed in A-E selection criteria. Agencies should also take steps to ensure that design fees include adequate money for estimating and that money earmarked for estimating is actu- ally used for Hat purpose.

OCR for page 19
SUGGESTIONS FOR IMPROVING THE EMMY ESTIMATING PR'EDURES USE OF PARAMETRIC ESTIMATES As discussed in the preceding chapter and AD pendix D, the Air Force and several private corpo- rations have developed parametric estimating sys- tems in which the estimated cost of constructing a proposed facility can be predicted by modifying the known cost of constructing a similar facility in the past on the basis of certain key parameters. Some committee members feel that parametric estimating techniques have advantages over tradi- tional estimating techniques, especially for early estimates. These members believe that traditional estimates are prone to error because they depend on the identification of every possible cost item in a project and the assignment of accurate cost values to those items. Since the likelihood of an item being missed or of a cost being incorrectly esti- mated is high, proponents of paramedic estimating believe it is better to start with a complete and accurate estimate and merely adjust to reflect the situation win regard to the proposed facility. Other committee members are skeptical of the claims made for parametric estimating. They feel that the validity of the formulae used to adjust costs on the basis of parameters has not yet been suffi- ciently validated to justify the widespread use of the techniques. These members tend to support the decision of Congress not to accept parametric esti- mates as a basis for requests for funding. (lIow- ever, Congress has authorized further investigation of the technique.) On the other hand, Air Force personnel have investigated He accuracy of their parametric esd- mating system and they believe that the system produces estimates that are at least as accurate as estimates prepared using more traditional methods (see Bridges and Gregory, 1988~. The consensus of the committee is that paramet- ric estimating is most appropriate when the facility being estimated is similar to a building for which cost data is available. The committee believes that pararne~ic estimating is sufficiently promising that 21 all agencies should consider using the technique more extensively. USE OF PROBABILISTIC ESTIMATING Periodically over the past 15 years it has been suggested that agencies could benefit from greater use of probabilistic estimating techniques such as range estimating (see Consulting Committee on Cost Engineering, 1983~. Range estimating, as discussed in Appendix D, is a technique that permits an esti- mator to quantify his confidence/uncertainty about an estimate. There is of course a degree of uncer- tainty about all estimates, but the uncertainty is highest early in the design process, particularly in unusual projects and projects with many unknowns. With probabilistic estimates, a number of pos- sible construction costs for a project are determined, and the probability of each value being the low bid is calculated. If an estimator's confidence in his data is high, the range or spread of possible costs will be narrow and the probabilities attached to the values near the mean will be high. Conversely, if an estimator's confidence in his data is low, the range of possible costs will be wide and the proba- bilities attached to values near the mean will be low. As in the case of parametric estimating, He commit members were split on the value of proba- bilistic estimates. Some members felt that proba- bilistic estimates more accurately reflect the real world, and thus are inherently better than single- value estimates. However, other members note that probabilistic estimates are more time-consum- ing to prepare than traditional estimates and that they can also be more difficult to interpret. The majority of committee members agreed that there is a benefit to viewing project budgets in terms of probability and risk. It is the committee's recommendation, therefore, that agencies consider preparing early estimates as probabilistic esti- mates~specially for unusual and high-risk proj- ects.

OCR for page 19