7 I would like to express my appreciation to Dr. Mendershausen for having called my attention to a number of errors which crept into my original calculations. I have now recalculated all figures for I958 and I959. The results are almost in substance although a number of the figures require minor corrections. By unchanged in substance, I mean (I) there is still mutual discrimination between the Soviets and the Bulgarians and (2) the Bulgarians still appear to discriminate more heavily than the Soviets against the rest of the Bloc, particularly with respect to exports. The only error of direction appears to be Bulgarian imports from the USSR in I959 (Table 2) in which it turns out that the Bulgarians did not discriminate against the Soviets on a by-commodity basis. However, they still discriminate in I958 on imports and also in I959 on a weighted value basis (Table 3). Recalculated figures are available upon request.

(3) Only one quarter of the estimates of al among 40 commodities and services, however, are significantly different from unity, which means that the tendency to commodity price equalization may be dominant. (4) Also, the possibility that price differences among countries may be explained by factors other than wages (such as prices of raw materials) appears to be high. If this possibility is sufficiently high, the tendency of commodity price equalization with respect to price of labor input will be strengthened, that is, estimates of al will approach unity under proper adjustments of raw material prices. This conjecture was tested for only two commodities, cotton shirts and blankets. In these cases, cost components of raw materials, namely raw cotton and raw wool, were taken out using the United States cost structure. After this adjustment, the following estimates were obtained: Before Adjustment After Adjustment Cotton shirts .9II -.936 Blankets -1I.I20 -.95o In conclusion, it should be mentioned that more proper selection of wage data (in this case the average wage in the manufacturing industry was used in every regression), addition of raw material prices, and of capital cost, would provide further interpretations.