This monograph articulates John Taylor's contributions to macroeconometric policy evaluation and design since the late 1970s. Yet, the book is not simply a collection of his previously published articles; instead, it is a well-organized premier for the new-Keynesian approach to macroeconomics. The author attempts to merge the fundamentals of the Keynesian tradition in modeling with the theoretical developments since the rational expectations revolution. He accomplishes this objective by demonstrating the feasibility of using a prototype model tractable by individual researchers, as opposed to large-scale models, to deliver scientific answers to practical questions in macroeconomic policymaking. Taylor's pedagogy involves building three frameworks with varying degrees of complexity and computational difficulty to reflect the issues addressed. In search of a world-economy model in Chapter 3, he progressively develops a preliminary stylized two-country model and then an intermediate VARMA model of five linear equations for the U.S. economy. For each model, he carefully demonstrates model construction, its estimation technique, simulation results, and provides policy analysis. Such treatments highlight the significance of different modeling strategies: modeling and computational costs versus econometric performance. For instance, while the VAR estimation method is feasible of estimating the closed-economy model, the multicountry model can only be estimated using the single equation FIML method. Nevertheless, even Taylor's workhorse-the world-economy model with ninety-eight stochastic equations for the G-7 countries-is still conceptually manageable. The models are prototypical of the new Keynesian approach. Their main features consist of rational expectations, staggered wageand price-determination mechanisms, and forward-looking consumer behavior. This framework contributes to the dynamics of long-run neutrality (monetary policy) as well as complete crowding out (fiscal policy) and persistent short-run fluctuations with a trade-off between the variances of inflation and real output. The treatments in the international sector include the Mundell/Flemming theory of perfect capital mobility as well as time-varying risk premia in foreign exchange and capital markets. In line with the recent developments in the time consistency and policy credibility theories, Taylor redefines policy rule as a well-defined contingency plan in a dynamic optimization problem [p. 5], rather than the constant money growth rule a la Friedman. In this sense, government policy follows a systematic feedback rule and is no longer exogenous. For policy evaluation, Taylor focuses on the variability of output and prices, rather than their levels as customarily considered in the literature. Estimation of the multicountry model reveals varying extents of wage and price synchronizations across different countries, with the highest in Japan and the lowest in the U.S. Such cross-sectional observations play an important role in explaining their different degrees of cyclical fluctuations and in determining optimal domestic policy rules. Moreover, since policy effects vary across countries, policy evaluation can not be addressed theoretically. Compared with other large-scale international models, such as LINK, the Federal Reserve's MCM, and the International Money Fund's MULTIMOD, Taylor's model is parsimony yet capable of delivering answers to many current issues. The entire Chapter 4 is devoted to the analysis of the world-economy model's structural residuals. The variance-covariance matrix of the shocks not only shows their different volatilities over different sectors and different countries, but also a high degree of correlation among these shocks. However, the model is estimated over the period 1972-86, which raises the issue of robustness in Taylor's conclusions. First, the global flexible exchange rate regime began in the early 1970s, rendering it inappropriate to analyze fixed versus flexible exchange rates in the following chapter. Second, structural changes in some of the equations are likely to occur over this turbulent period. Examples for the U.S. include changes in the wageand pricesetting mechanisms and in the monetary policy regime, and deregulation in the financial markets within that period. In Chapter 5, Taylor compares the effects of unanticipated versus anticipated monetary and fiscal policies. The anticipated policy is simulated with a two-year learning period between a policy announcement and actual policy implementation. In the short run, unanticipated policies result in higher volatility in real output and prices than anticipated actions. Policy effects are also compared over the flexible and fixed exchange rate systems. Transmission of policy effects abroad is more significant under fixed exchange rates when money supplies and interest rates are manipulated to keep exchange rates pegged. Chapters 6 and 7 on policy design and implementation, respectively, form the core of the book. In policy design, the author sequentially evaluates the exchange rate systems, the prospect of international