How much did prices fall?

In my view most results in chapter 4 should not arouse controversy. The results for defence industry value added deserve close scrutiny, however, given what they imply for defence industry employment and productivity, and their influence on the evaluation of industry as a whole. The suggested 2.2-fold increase over 1940–4 in value added per hour worked in Soviet defence industry, and 3-fold increase in value added per worker, are certainly large enough to evoke surprise. Are such figures plausible?

The reader should not underestimate the importance of this issue. Present estimates of defence industry value added, employment and productivity are based on a complex structure of data, assumptions, and reasoning. This may make for an appearance of fragility. Results are obtained which make a startling impression. Their plausibility may be questioned – yet, at the same time, they are very hard to undermine on the basis of contemporaneous data and comparative trends. If these estimates are badly wrong, however, then a wide range of basic data consistent with them must also be called into question. If average costs fell dramatically, and labour requirements did not fall, material costs must have fallen still more dramatically; or, if material costs did not fall, labour requirements must have fallen by still more. If average costs did not fall, the trend of defence industry production must have been greatly exaggerated, or the evidence of price change must be in some way or other highly misleading, or our understanding of the budgetary record, and the mechanics of the defence procurement process, must be fraught with misinterpretation.