VI

CONCLUSION

The analysis of the spatial efficiency of the Soviet cement industry has yielded several results that might have been expected and a few that are surprising. Almost all of them may be expressed quantitatively and their relative importance established.

A. SHORT-RUN PROBLEMS

The Soviet approach to product standardization and performance indicators was the major short-run source of waste to the economy. Our analysis showed that there would have been a small saving in production costs and a large saving in transport costs with a standards policy like that in the United States industry, which would have promoted the production of higher-quality cements. Instead, the industry was geared to a success indicator expressed in tons which encouraged the production of lower-strength cements. Had the specifications of grade 400 been set as the minimum acceptable standards for Portland cement in the 1930’s, overexpenditures of 28 percent of transport cost actually incurred, 5.5 percent in production costs, and 17 percent in total delivered costs would have been avoided. Instead, only 20 percent of cement production (and that only late in the decade) met the requirements of Portland grade 400 or higher, although there was no technological obstacle to production of cement of this quality. Since the problem was so significant for this industry, it would seem to be of even greater consequence in other industries where standardization of the product mix and performance measurement are more difficult.

The second important short-run conclusion concerns the efficiency of the distribution of the cement that actually was produced. During the first three five-year plans, which spanned the years from 1928 until World War II, when the economy was experimenting with new forms and methods and was adapting to the realities and difficulties involved in instituting socialism, the shipment patterns adopted by the industry to link producers and consumers were remarkably efficient. Our optimal linear program solution for 1936 revealed an overexpenditure of transport inputs of no more than 7 percent. The conclusion about rational distribution is further supported by analyses of 1935 and the water-route use pattern. The observed efficiency contrasts with the protests of numerous critics of the industry’s marketing and shipping practices of the time. How was this achieved?

The success shown in marketing and shipping operations was the result of good planning by the central marketing organization. As explained in Chapter III, centralized order-taking and product distribution was the one source of stability in the industry’s many reorganizations during the period. Even before the development of linear programming techniques, this marketing organization had mastered the distributional skills needed to ship the industry’s output with very little waste. It did this in spite of the pre-1939 rail-rate structure, which by its rapidly declining charge per kilometer encouraged long hauls and in spite of the cement-pricing policies of the 1930’s in which the buyer had to pay all freight costs regardless of the origin from which the marketing agency chose to supply him.

Can the results of our analysis of cement industry operations be extended to freight transportation generally? That is to say, did irrational shipping operations typically account for such a small proportion of recorded ton-mileage? One is tempted to believe that what we found for cement in this respect may also hold for other relatively homogeneous commodities such as coal, firewood, lumber, petroleum, and grain, which comprised about half of the originated tonnage during the 1930’s and whose distribution plans should not be appreciably more difficult than that for cement. For some of these commodities, however—firewood for example—the rate structure may have been a more important problem, encouraging high consumption through an unrealistically low transport cost, and in this way leading to an excess of the transport input. This contrasts with cement demand, which all considerations led us to conclude was inelastic. With regard to irrationality among less homogeneous commodity groups, such as iron and steel products, there was probably what Hunter has called a “looser form of crosshauling” in which owing to plant specialization and its resulting economies, one item in the product group went in one direction at the same time that a second went the other way.¹ When the product group is looked at as a whole, there would appear to be crosshauling. Whether it is irrational would depend on the magnitude of the scale economies in question.

B. LOCATION

Our assessment of long-run efficiency is less flattering—the location decisions made by Soviet planners for the cement industry could have been much better than they were. Over the decade the regional imbalances between production and consumption worsened, as is shown most graphically in the increase in the average haul of cement over the decade. The most notable deficiency in planning of new capacity appeared in the failure to build more in the eastern regions. The analysis of 1940 in Chapter IV showed that an important saving in transport would have resulted from a better planned distribution of new capacity providing more production in the east. Through better locational decisions, total delivered cost could have fallen by 33 percent. This means that total delivered costs were 50 percent greater than they would have been in conditions of economic efficiency.

Most of the unjustified decisions regarding new cement capacity during this period reflect a common failing—they emphasized production costs to the neglect of total delivered costs. This was promoted by the nature of the industry’s success indicators; since total output was the indicator of greatest consequence, the industry naturally had an incentive to produce as much as possible with the funds at its disposal. This policy would naturally lead to production in the West, where production costs were lower. Perhaps if the rail-rate principle of constant incremental charges per kilometer from around 500 kilometers upward, which took effect in 1939, had been operative earlier, planners would have been more aware of the high costs of transport, and this might have brought home the need to locate production closer to consumption, which, thanks to the relative absence of technological constraints, would have been a practicable policy. And it should be emphasized that even the gains in mill costs secured by following the actual policy of producing in the West were only 7 percent of the production costs that would have prevailed under the optimal location variant.

The buildup in the Ukraine—based on lower production costs there than elsewhere—and the relative neglect of the East constituted the major factor in the 1930’s in the increase in overexpenditure of resources by the industry. A less important factor was the neglect of small plants based on a different technology. It is not so much that the often censured gigantomania of Soviet planners of the period afflicted cement industry planners—the average size of plants put into operation during the period was 100,000 to 120,000 tons; and of the initial 1931 plans calling for eight portland or portland-slag plants with capacities in the 232,000 to 465,000-ton range, all but one were ultimately disregarded, the one undertaken being in the Moscow area which was the nation’s largest consuming region. But the possibilities of the small 40,000-ton plant based on shaft kilns do seem to have been overlooked, although there were some areas where such plants could have been an efficient solution.

It may be of interest at this point to observe that since 1940 there has been a great increase in eastern production—including Eastern Siberia, where production rose from zero in 1940 to almost 1.4 million tons in 1958. This has helped increase self-sufficiency in the Far East and Eastern Siberia from 44 percent in 1940 to 90 percent in 1958, and has been an important factor in reducing the average length of haul to 600 kilometers in 1958.²

Finally, we must ask to what extent the inefficiency uncovered here was characteristic of industry generally. It may be that the location planning of other industries was worse for the reasons that the demand for the output of most other industries is more difficult to project and their cost functions are more difficult to determine. The failure to take proper account of the relationship between production costs and transport costs in the cement industry, which was not marked by great regional or scale cost variation, may mean that in industries with more complex demand and cost structures this relationship was less successfully dealt with, causing greater overexpenditure of transport per ton of product than in the cement industry.

If our conclusion about transport use in the cement industry is valid for the economy as a whole, a very large part of the more than fourfold growth observed in transport output during the first three Five Year Plans reflected inefficient location patterns. The 70-percent transport overinput arising in faulty cement industry location involved a net resource overexpenditure of around 40 percent in delivered costs. For industries producing commodities with higher value per ton than the cement industry, the transport overexpenditure would have been less important when related to total delivered costs even while the physical input averages may have been the same. If there are enough such sectors, naturally, they will affect the net balance somewhat, so that the total national waste arising from poor location may be less than we have observed in the cement industry.