Monetary and Public Debt Shocks: Tests and Efficient Estimates
用有效信息最大似然法重新检验Barro关于边际税率跨期效率的假设，直接检验理性预期和结构中性命题，对研究财政赤字与货币政策关系的学者有参考价值。
IN AN ARTICLE IN THE JMCB, Barro [1] defined a form of intertemporal efficiency for the behavior of marginal tax rates (and expected tax rates). Given certain additional conditions on government expenditures, Barro's definition of efficiency implied that particular relationships and time paths would characterize federal deficits. In that paper, two of the seven predicted relationships were tested: (1) the sign of the relationship between the deficit and the ratio of output to its trend value, and (2) the effect of anticipated inflation on the deficit. Those tests were performed under the joint assumptions of the rational expectations hypothesis (REH) and predictable policy irrelevance (or structural neutrality (SN)) with regard to both monetary and debt policies. In this paper, Barro's tests of the efficiency hypothesis are reexamined within a wider context by means of the statistically efficient full information maximum likelihood (FIML) method [4 6].1 Specifically, by allowing the imposition of cross-equation and within-equation parameter constraints, the FIML method makes it possible to: (a) directly test the REH and SN propositions (jointly and separately