The Evolution of Spending Multiplier Economics: From Neoclassical to New Keynesian and Beyond
系统回顾1980年代以来政府支出乘数文献，发现乘数大小因经济条件、识别方法和模型假设而异，不存在统一估计，尤其在衰退、零利率下限或固定汇率下乘数更大。
ABSTRACT This paper provides a comprehensive and systematic review of the literature on government spending multipliers from the 1980s to the present. Contributions are organized along theoretical and empirical dimensions, revealing a highly heterogeneous and context‐dependent landscape. The evidence consistently challenges the notion of a universal or fixed “Keynesian Multiplier”: instead, multiplier size varies significantly depending on economic conditions, identification methods, and model assumptions. While traditional real business cycle and representative‐agent new Keynesian models fail to generate multipliers above unity, recent advances—particularly those involving heterogeneous agents or behavioral frictions—yield more substantial effects. Empirically, results are highly sensitive to the treatment of fiscal shocks and the underlying identification strategy. Moreover, state dependence emerges as a key theme, with larger multipliers often observed during recessions, at the zero lower bound, or under fixed exchange rate regimes. The size and effectiveness of fiscal multipliers depend on the specific economic environment, modelling approach, and empirical framework. There is no “one‐size‐fits‐all” estimate: instead, multipliers are inherently conditional and vary across time, space, and institutional context.