Here’s Admiral Hyman Rickover in a 1971 Appropriations hearing*:
There is an essential difference between the nontechnically trained or experienced administrator and the engineer. Administration is necessarily based on the law of averages. The pure administrator learns how people will act “on the average” and he makes decisions accordingly. Therefore, he can be promoted to ever higher positions and continue to use the “law of averages.”
The engineer, on the other hand, cannot be governed by statistical averages. Each decision he makes is concerned with a specific item. That item must work. It isn’t enough that such items will work “on the average.” Therefore, the effective engineer, if he is to do an adequate job, is condemned to being concerned with details. A single apparently minor detail can wreck a major project even though all the other parts work. This constant attention to detail is a prerequisite — the hallmark of an engineer worth his salt. The engineer’s product either works or it is “junk.”
The whole tendency of the Navy is to do away with technical expertise, and to have the remaining people become “managers.”
Obviously this is true of cost estimation for new programs. Estimators simply go to cost histories, run a regression on cost vs. weight (or some abstracted measure of complexity/capability), and spit out an “average” number. Then, when the program proceeds, earned value management data gives you the “average” cost performance index across the contract.
This quote reminded me of Frederick Mosher who made a similar comment about the programming of the budget, where again we have generalizing administrators take over the operating plans of highly complex organizations that depend on specifics:
There is one other kind of inconsistency. It is the mixture of operational functions, involving tremendous bodies of detailed and specific information, with generalizing, thinking, and planning functions such as those involved in management and budget planning, in accounting policy development, and in program review and interpretation at the departmental level. It is almost an axiom of administrative organization that the former kinds of activity, when placed in the same organizational setting as the latter, tend to reduce them to their own level. A major need in the budget planning and programming systems in the military and, in fact, in the entire government, is to escape the mass of operational detail, to provide tools and brains for generalizing and synthesizing.
Budgeting and acquisition administration are not neutral to program outcomes. The more layers of management and control are put onto the budget structure, the more impossible it becomes for technical experts to execute a successful program.
*DEPARTMENT OF DEFENSE APPROPRIATIONS FOR 1971. HEARINGS BEFORE A SUBCOMMITTEE OF THE COMMITTEE ON APPROPRIATIONS HOUSE OF REPRESENTATIVES NINEITY-FIRST CONGRESS SECOND SESSION. PART 7. Testimony of Vice Adm. Hyman G. Rickover.
Leave a Reply