Significant annual cost impacts for the Federal civilian agencies in the post-transition period of a measurement-language-only change were estimated as follows: the Department of Agriculture-$3 million savings, mostly in the Agricultural Research Service; the National Aeronautics and Space Administration- $445,000 savings (incomplete figure); the U.S. Postal Service-$250,000 added costs, in the Letter Mail Equipment Branch (development costs); the Food and Drug Administration - $200,000 savings; the Tennessee Valley Authority-$200,000 savings, mostly in the Office of Power (in engineering activities), the Office of Engineering Design and Construction and the Maps and Surveys Branch; the U.S. Tariff Commission $187,000 added costs (attributable to statistical analyses, investigations and research); the Small Business Administration - $156,000 savings; the U.S. Coast Guard - $132,000 added costs, in the Office of Engineering (mainly for conversion of specs, aids to navigation activities, and training); and the Atomic Energy Commission-$132,000 added costs, savings in the Divisions of Space Nuclear Systems and Isotopes Development falling short of the expected costs in the Naval Reactors Division.

Anticipated Impacts of a Planned National Program to Increase the Use of Metric Measurements Units and Engineering Standards - Under the assumption of a 10-year coordinated national program to increase the use of metric measurement units and metric engineering standards (Assumption III), the Federal civilian agencies were asked the same set of questions as under Assumption II above. (See app. 3, pp. 54 and 62 for detailed definition of the Assumption.) In a gross sense, responses under Assumption III paralleled those under Assumption II. The measurement-language change in and of itself would have more impact on most activities than the revision of engineering standards to a rational metric base by itself (which is the added factor in Assumption III). Of the 394 subunit respondents, at least 18 percent are entirely uninvolved with engineering standards in their work activities, although they use measurements. (These respondents all answered "don't know" or left blank the questions under Assumption III.)

All of the advantages listed under Assumption II (adoption of metric measurement units only) were also listed for metrication under Assumption III, since the latter includes adoption of metric measurement units. Additional advantages deriving from the adjustment of engineering standards to a rational metric base would include: facilitated international promotion of U.S. standards, reduction of dual standards in international agreements, greatly facilitated relations of the production community to the scientific community, and the possibility of adoption of a more realistic system of nominal sizes and subdivisions thereof. The Bonneville Power Administration in the Department of the Interior and the Tennessee Valley Authority are representative of hardware-using agencies which foresee significant advantages from metrication including engineering standards in the potential for increased harmonization of standards internationally, permitting greater interchangeability of materials, parts and equipment. International comparability of articles, which would improve somewhat with the adoption of metric measurement units according to the U.S. Tariff Commission, would be enhanced even more by eventual harmonization of standards on a metric base.

As under Assumption II, the primary disadvantages listed were operational impairment and cost increases related to the operation and maintenance of equipment and structures designed to customary dimensions and engineering standards. These problems would be somewhat more serious in this case, since manufacturers would eventually begin switching to the production of equipment and parts designed to the revised, metric-based standards, and the maintenance of replacement parts for older, customaryengineered equipment might become somewhat more difficult and costly.

In this context, six respondents who feel that long-term advantages of metrication of measurement units would outweigh any disadvantages, hold the opposite view with regard to metrication to include engineering standards. They are: Facilities Management in the Internal Revenue Service, State and Private Forestry in the Department of Agriculture, the Weather Bureau in ESSA, the Bureau of Abatement and Control in the Environmental Health Service (Department of HEW), the Office of Automated Data Management Services in the General Services Administration, and the Lewis Research Center of NASA. On the other hand, five respondents hold just the opposite split of views, believing that metrication of units only would be disadvantageous, on balance, and metrication including revision of engineering standards advantageous. These are: Administration in the Forest Service (Department of Agriculture), the Kansas City Division of Bendix Corporation (AEC contractor), the Office of Design and Construction in the General Services Administration, the Letter Mail Equipment Branch in the U.S. Postal Service, and the Office of Construction in the Veterans Administration. The Building Research Division in the National Bureau of Standards does not know whether advantages would predominate under Assumption II, but feels they would under Assumption III. As the Forest Service Office of Administration put it, "Metrication in units would not take care of the real problem. The real problem would be solved through metrication of both language and hardware." (Obviously at some point in future time, whether in 20, 30, or 40 years, all equipment will have been replaced for reasons of wearout or obsolescence-40 years is probably an excessive upper bound, in view of the pace of technological change today - and there would then be no compatibility problem. The question is just how much difficulty would be experienced during the 20- to 40-year period.)

All told, 207 of the 394 respondents (60% of those answering the question) expect that long-term advantages would predominate with metrication under Assumption III, and 47 (14% of those answering) think they would not. (As mentioned above, a number of respondents did not answer questions under this Assumption since they are in no way involved with engineering standards in their work.) As under Assumption II, the only agencies with pluralities of respondents holding that long-term advantages would not predominate under this Assumption are the U.S. Travel Service and the Federal Trade Commission. Three agencies have greater pluralities of responding units which expect advantages to predominate under Assumption III than with metrication of measurement units only: the National Bureau of Standards, the U.S. Postal Service, and the Atomic Energy Commission. Of the 18 agencies with no negative responses as to the predominance

of advantages under Assumption II, 15 hold the same view with regard to Assumption III. The exceptions are the Environmental Health Service (Department of HEW), NASA, and the Treasury Department (each of which had one respondent holding the contrary opinion under Assumption III). The Patent Office observed that, even though the advantages are intangible, the opportunity to increase compatibility and standardization on the international level would be significant. In spite of the particular difficulties and efforts that would be encountered in the hardware-involving, operational activities of the Atomic Energy Commission and NASA, most reporting groups in both of these agencies feel that, at least in non-fiscal terms, advantages of metrication would predominate in the long run. Some respondents in these two organizations even foresee eventual fiscal dividends. The general view in AEC and NASA regarding Assumption II versus Assumption III seems to be that, in the long term there would be more gained from a metric conversion that includes revision of engineering standards as well as units. The Veterans Administration feels that "in both health facilities construction and health services operations there would be advantages in the improvement of international communications and in the promotion of U.S. standards. We believe that the advantages of adopting the metric measurement units and engineering standards outweigh any disadvantages, and would be worth the cost as far as our internal operations in the health area are concerned." The Tennessee Valley Authority stated that "advantages of the changeover to the metric system would far outweigh the disadvantages." The problems of a purely measurement nature attendant to metrication under Assumption III would be the same as those under the previous Assumption. Additional difficulties raised by the eventual revision of engineering standards would include: physical adjustments in building construction, space layout and procurement functions; equipment maintenance and servicing (as discussed above under disadvantages); and acquisition of some additional tools. The AEC observed that these problems would require careful training and supervision of craftsmen, and could lead to errors of interpretation. Several subdivisions in AEC believe they would have to make substantial investments in purchasing new or modifying existing equipment and instruments, and some dual inventories would have to be carried. The Commission feels it unlikely that the conversion of codes and standards to an SI base could be accomplished without specific subsidization by the government, "since the country is already in some difficulties in keeping codes in step with modern materials and processes."

The Federal Power Commission (FPC) stated that the revision of engineering standards would affect its responsibility in studying the safety and adequacy of licensed hydroelectric projects, since this activity involves safety codes covering electrical, structural and hydroelectric design criteria and employing customary measurement terms. However, "the difficulty of performing engineering analysis under two sets of standards would be essentially one of familiarizing staff with metric engineering methods." In the General Services Administration, in addition to the implementing activities mentioned under Assumption II, the Quality Control Division and the Material Evaluation and Development Laboratory would be involved in

reissuing many specifications and standards. NASA would face hardwaretype changes similar to those in the AEC. The U.S. Postal Service would have to revise engineering specifications for its equipment, and the Government Printing Office (GPO) mentioned conversion or duplication of some of its shop equipment. In the area of legal problems, GPO cited changes in contractual documents concerning machinery and equipment, including warranties. (See app. 6 for listing of legal problems.)

All of the general comments made earlier about difficulties of estimating the cost impact of metrication apply equally to the figures reported under Assumption III. For most agencies which expect significant costs attendant to the measurement-language change, estimates of transition costs run somewhat higher for metrication including revision of engineering standards. (Set against this increased cost is the very difficult to evaluate benefit that, having been revised in the light of up-to-date technological knowledge, our engineering standards would be, in the aggregate, better at the end of the transition than they are today.) Net totals of Federal civilian agency cost/savings estimates for a planned, national metrication program including revision of engineering standards to a rational metric base come to: (1) an average $58 million per year added cost during the nominal 10-year transition period and (2) an annual $7.4 million cost savings thereafter.

The 21 agencies which would expect no significant transition cost impact under Assumption II forecast a similar impact under this Assumption. The Veterans Administration, which foresaw no net cost impact under the earlier assumption, estimated an average $20,000 annual savings under this assumption. The Food and Drug Administration, which would expect $100,000 annual savings during a measurement-language conversion, estimated $60,000 annual savings if engineering standards are also revised.

The same nine agencies which would expect $1 million or greater annual transition costs under Assumption II comprise that category under this Assumption: The U.S. Coast Guard - $15.2 million, in the Offices of Operations and Engineering; the Atomic Energy Commission-$9.9 million, mostly in the Naval Reactors Division, the Division of Reactor Development and Technology (for retraining and recalibrating), the Mound Laboratory of Monsanto Research Corporation (primarily for production activities), the Nevada Operations Office (engineering, design, construction and maintenance activities), and the Space Nuclear Systems Division (the Division of Isotopes Development expects a $500,000 average annual savings during transition); the Department of Agriculture-$8.7 million, mostly in the Soil Conservation Service, the Agricultural Research Service, and the Extension Service; the Federal Highway Administration - $7 million, in the Bureau of Public Roads (reflecting increased design costs); the U.S. Postal Service-$3.8 million, mostly in the Building Design Division and for the adaptation or replacement of scales; the Department of the Interior - $3.1 million, almost entirely in the Geological Survey (the Bonneville Power Administration again expects costs and savings to balance out over the 10-year transition); the General Services Administration - $2 million (the increase over the Assumption II figure is in the Federal Supply Service, mainly for inventories and specifications); the Federal Aviation Administration - $1.7

million, mainly in the Systems Research and Development Service; and the Tennessee Valley Authority-$1.3 million, the additions being largely in the Office of Power (for revising internal specifications, and operations and maintenance including inventory) and the Office of Engineering Design and Construction.

Other agencies forecasting significant annual transition cost impacts under Assumption III are: the National Aeronautics and Space Administration- $849,000 (incomplete figure); the National Bureau of Standards $767,000; the Environmental Science Services Administration (now in NOAA)-$730,000; the Health Services and Mental Health Administration-$708,000; the U.S. Tariff Commission- $575,000; the Government Printing Office-$385,000; and the Department of Housing and Urban Development -$253,000 (incomplete estimate).

Significant annual cost impacts expected by the Federal civilian agencies after a 10-year metrication period under Assumption III are as follows: the Federal Highway Administration-$5 million savings, in the Bureau of Public Roads (from decreased design costs); the Department of Agriculture - $3 million savings, largely in the Agricultural Research Service; the U.S. Coast Guard-$2.2 million added costs, in the Office of Engineering (mainly for "custom manufacture of parts to engineering systems"); the Atomic Energy Commission - $458,000 savings (incomplete figure), savings in the Divisions of Space Nuclear Systems and Isotopes Development, and the Sandia Labs exceeding expected added costs in the Naval Reactors Division; the Health Services and Mental Health Administration $322,000 savings; the Veterans Administration - $240,000 savings, in the Office of Construction; the Food and Drug Administration- $220,000 savings; the U.S. Tariff Commission- $166,000 added costs, largely for investigations and statistical analyses; the Small Business Administration $156,000 savings; and the Government Printing Office-$137,000 savings.

What Action Should Be Taken- The final pair of questions in this Survey asked whether the respondents favored concerted action to bring about metrication in the United States, under each of the Assumptions, II and III. The agencies were also requested to provide an overall viewpoint on these questions in their agency summaries (see "Nature of the Survey," p. 8). Of the 50 agencies surveyed, 39 either expressed an overall view, or had a majority of respondents, in favor of metrication under both Assumptions. (See app. 4.) Three agencies had pluralities of respondents in favor of a U.S. metrication effort: the Federal Maritime Commission, the General Services Administration and the Government Printing Office. The Federal Maritime Commission observed that it is uninvolved with engineering standards, and none of its respondents expressed an opinion on metrication under Assumption III. The National Institutes of Health (in HEW) submitted a bifurcated view-one respondent in favor and one opposed - not resolved at the agency level. In one agency, the Social Security Administration in HEW, a majority of the five respondents oppose U.S. metrication (although none of the five foresees any cost impacts, none expects any disadvantages, and only one anticipates any problems during a transition). The remaining six agencies