The reason for the apparently opposing results in Brada and Graves' (1988) attempt to explain the reasons for the slowdown in USSR defense expenditures in the mid-1970s is that their analysis suffers from a serious serial correlation problem. The majority of the regressions display Durbin-Watson statistics that reject the null hypothesis of no autocorrelation. A reestimation of their results, after correcting for serial correlation, changes some of their major conclusions regarding the factors influencing Soviet defense spending. The corrected results indicate that no structural break occurred in the mid-1970s. These results suggest that there has been no change in Soviet military doctrine or in the Soviet leadership's preferences in the 1970s. In reply, Brada and Graves argue that the evidence for the existence of serially correlated disturbances is much more tenuous than Chowdhury suggests and that the evidence is more consistent with the existence of a structural break and no serial correlation of disturbances.