Does Anticipated Monetary Policy Matter?: Another Look
质疑Gordon和Mishkin关于预期货币政策影响产出的研究，指出其假设产出为确定性趋势平稳波动，但实际可能是非平稳过程，导致检验结果不可靠。
RECENT WORK BY Mishkin (1982a, 1982b) and Gordon (1982) tests Lucas-Sargent-Wallace (henceforth LSW) policy ineffectiveness proposition with econometric methods. This work generally finds support for alternative proposition that there is some short-term influence of anticipated demand policy or anticipated money growth on output. As McCallum (1982) has pointed out, one difficulty with these studies is they presume movements in output and employment are best represented as stationary fluctuations around deterministic trends, in constrast to evidence given by Nelson and Plosser (1982), which suggests these variables are nonstationary processes that have no tendency to return to a deterministic path. If these series are nonstationary processes, then distribution theory implicit in hypothesis tests of Gordon and Mishkin studies does not apply and their results are open to question. Also, as McCallum (1982, p. 4) points out, the methods of overcoming observational equivalence difficulty (Sargent 1976) are not entirely satisfying in these studies. In effect, Gordon and Mishkin are forced to make some untested restrictions in order to distinguish anticipated from unanticipated components