Are the Soviets Really Outspending the U.S. on Defense?

A r e the Soviets Really Outspending the U.S. on1 Holzman D. I T h e most widely used comparisons of the US and Soviet military expenditures are those presented by the U.S. Central Intelligence Agency. These comparisons have serious deficienciesand specificbiases. The purposes of this articleare twofold. First, several sources of bias, the result of which in each case is to exaggerate Soviet defense expenditures, should be exposed. In addition, the implications of these flawed estimates for the national security debate are critical, and must equally be brought to light. Let me state at the outset that the two most obvious sources of bias are the failures to take account of the relatively higher quality of both U.S. military personnel and U.S. military equipment. Further suspicion of bias stems from the fact that so-called ”index number effects” in the CIA estimates are so much smaller than those usually experienced in U.S .-Soviet comparisons. Some crude attempts are made here to pin down the orders of magnitude of bias in the CIA estimates. These efforts were handicapped not only by unavoidable methodological and data problems but also by the paucity of information underlying the CIA estimates which is released by that agency. 1. After this paper was completed, I received a copy of: U. S. Arms Control and Disarmament Agency, World Military Expenditures and Arms Transfers, 1968-1977, Washington, Oct. 1979. In a short essay entitled “Soviet Military Expenditures,” the Agency expresses its dissatisfaction with the CIA figures and presents some of the arguments which are dealt with below. Their overall view of CIA estimates is: ’ I . . . estimates of this type probably overstate the relative size of Soviet military expenditures compared to the military spending of the United States . . .these estimates of Soviet military spending may not be the best answer to the question: What single valuation of Soviet military spending yields an evenhanded comparison with U.S. spending” (p. 13). The Agency promises further research on alternative valuations (p. 15). Readers are referred to their excellent statement of issues. It is unfortunate, in light of its dissatisfaction with CIA estimates, that the Agency otherwise continues to use the CIA figures as the best available in U.S.-Soviet military comparisons. The inspiration to write this paper came from attending a meeting at the Stockholm International Peace Research Institute. A first draft was completed while I was a research fellow at the Atlantic Institute in Pans, where I benefited from discussions with Gregory Flynn, Martin Hillenbrand, and Robert Lieber. Later versions benefited from discussions with my colleague at Tufts, David Dapice, and from a presentation at the Harvard Russian Research Center in late 1979, where the comments of Abram Bergson, Edward Hewett, and Barney Schwalberg were particularly helpful. Franklyn D. Holman is Professor of Economics at Tufts University, and an Associate of the Russian Research Center, Haroard University. 86 Are the Soviets Outspending the U.S.? 1 87 Let me stress also that while the volume of military expenditures is taken here as a proxy for military capability, it is not a very good one. Military expenditures represent the costs of inputs and the performances (outputs) from these inputs are not necessarily proportional to the costs. All this is admitted by the CIA. It uses the value of military expenditures in lieu of a better aggregative measure of military capability. For purposes of this paper, the CIA’Sexpedient is accepted without question.* The lndex Number Effect The CIA regularly publishes estimates of aggregate Soviet defense expenditures in dollars, estimates which are then compared with the defense expenditure total presented in the U.S. federal budget. The CIA estimate is necessitated by the facts that 1)the USSR publishes only one figure for defense and it is commonly agreed that that figure substantially understates the Soviet defense effort; 2) even if a comprehensive figure for defense were published, that figure could not be easily used for purposes of comparison since the ruble exchange rate is not a real price. The CIA procedure, in brief, is to try to get quantity estimates of everything that is included in defensenumbers of soldiers, weapons, etc.-and to value...