中文摘要：

论文研究了中国通胀的学习型预期，以产出缺口为驱动因素的中国通胀动态模型。并对其进行了扩展，利用递归预测方式，基于VAR和BVAR模型遴选出货币、成本、国际、资产价格因素的代理指标，分别构建了学习型预期和消费者预期指数下中国扩展的新菲利普斯曲线。基于Granger因果关系、脉冲响应和方差分解验证了前面的实证结论。针对前面研究的不足，进一步基于Markov机制转换模型研究了中国通胀水平的驱动因素。分析表明：中国通胀惯性非常强，引入宏观经济变量几乎不能够降低通胀惯性；公众预期能力不强，消费者预期指数以及学习型预期对通胀的决定作用有限，后顾性因素比前瞻性因素在短期通胀动态中起更重要的作用；菲利普斯曲线起作用的持续时间长于货币数量论起作用的持续时间，在大多数时候，产出缺口相对于货币增长更有可能是通胀的驱动因素；各种方法都表明农业生产资料价格是通胀的重要驱动因素。

英文摘要：

The paper studies the learning expectation and the dynamic model of inflation driven by output gap in China. Chinese expanded new Phillips curves under learning expectation and consumer expectation index are constructed by selecting the proxy indicators for money, cost, international, asset price factors using rolling forecast based on VAR and BVAR models. The previous empirical conclusions are verified based on Granger causality test, impulse response and variance decomposition. To the lack of previous studies, the drivers for China＇s inflation are studied further based on the Markov switching model. Analyses show that： Chinese inflation inertia is very strong and almost cannot be reduced by introducing c variables; Public expecration ability is not strong, and the role of consumer expectation index and learning expectation on the inflation decision is limited and backward - looking factor play a more important role than forward - looking factor in the short - term inflation dynamics ; The duration of Phillips curve working is longer than the duration of the quantity theory of money working and in most cases the output gap is more likely than monetary growth to be the driving factor to inflation; Various methods have shown that the price of agricultural production is an important driving factor to inflation.