On the Reinterpretation of Money Demand Regressions
检验了Goodfriend（1985）关于货币需求回归中滞后因变量系数过大的解释，即自变量存在序列相关测量误差，并提出了检验方法。使用Goldfeld（1973）数据集的检验结果明确拒绝了该假说。
A stylized fact concerning estimated money demand relationships is that lagged dependent variables have high explanatory power and large estimated coefficients, which is hard to explain at a theoretical level. Marvin S. Goodfriend (1985) suggests that this may be due to the presence of serially correlated measurement errors in the independent variables. The author demonstrates how consistent estimates of the short-run demand function can be obtained even if the Goodfriend hypothesis is accepted and also how the hypothesis might be tested. Application of the suggested test using the Goldfeld (1973) data set leads to decisive rejection of the hypothesis. Copyright 1994 by Ohio State University Press.