Study On Internal-propagation Mechanism Of Stochastic Growth Model

In the last 30 years, dynamic stochastic general equilibrium (DSGE) model is the most popular optimized basic model in the macroeconomic research field, and also a new perspective for economical modeling, which makes plenty of scholars study on. Although classic econometrical models remain to be the main tools for analysis and forecasting, DSGE model reflects many unique characters by the clear description of the relationship among economic agents'behavior and the meticulous characterization between long-term economic equilibrium and short-term dynamic adjustment process, such as explicit modeling framework, theoretical consistency, prefect combination of macroeconomics and microeconomics analysis and organic integration of long-term and short-term analysis, which becomes the researcher's favorite. Meanwhile, the speedy development of IT and the improvement of the Bayesian methods also makes a deep impact on the evolution of DSGE model. DSGE model is becoming the benchmark model for many research institutions to do quantitative analysis.DSGE model is based on the behavior decisions of economic agents, and the behavior equation of total economy is obtained with appropriate aggregation technology, which fundamentally guarantees the consistency between macroeconomics and microeconomics analysis, makes the model possess an overall properties, and also integrate the long-term and short-term analysis. In addition, DSGE model can fully describe the setting and identification of economic agents behavior decisions in uncertain environment, and the structure character of DSGE model can keep the model away from Lucas critique, which ensures the model to play a more important role on policy analysis and evaluation, provides convenience to analysis framework and welfare analysis based on optimization and create the possibility for the optimized choice of policy and the comparison of various policies. These advantages cannot be owned by other economic models. Furthermore, the discovery of DSGE model makes the policy makers and theory researchers no longer be satisfied to give out numbers, but become more concentrated on the implied practical meanings under the numbers, and strive to explain them theoretically, which enrich the content of economic analysis and currency policy decision.In some developed countries, DSGE model has played an important role in economic analysis and currency policy decisions, even in economic forecasting. In the new century, the people's bank of China also imports the DSGE model. However, due to the early phase of the development, lying in the primary stage, DSGE model is far from completion, which causes all kinds of deficiencies. The lack of endogenous dynamic in the model is one of them, influencing the overall effect of model fitting, and eventually causing the accuracy of forecasting. Therefore, it is vital to solve this problem.In this paper, we are mainly against the problem presented above. We modify the stochastic growth model, use Bayesian approaches with GDP growth per capita and hours worked per capita data in Taiwan and South Korea, and give advices for the expansion of internal-propagation mechanism in two aspects as follows:Initially, we modify the stochastic growth model by incorporating non-stationary labor supply shocks, estimate the standard specification in which hours worked are stationary and a modified version with non-stationary labor supply shocks, and compare the fitting effects between models in different versions. We find that, if firms can freely adjust labor inputs, the data support the latter specification, which means the introductions of non-stationary labor supply shocks efficiently improves the internal-propagation mechanism of stochastic growth model. Once we introduce frictions in terms of labor adjustment costs, the overall time series fit improves and the model specification in which labor supply shocks and hours worked are stationary is preferred, which means the introduction of labor adjustment improves the internal-propagation mechanism of all model versions, and especially improves internal-propagation mechanism of the model version with stationary labor supply shocks.What's more, we modify the stochastic growth model by incorporating a"learning by doing (LBD)"mechanism, intending to provide the internal-propagation mechanism of DSGE model. We find that, though there is a possibility of response lag, the internal-propagation mechanism of real business cycle model to GDP and hours worked has been apparently improved by incorporating LBD mechanism.Combined with the empirical results in this paper, firstly, we suggest that we must strengthen the research on DSGE model in our country, which is vital to the making of macroeconomic policy. Studying on the DSGE model that is accessible to our country can provide an effective supplement and a scientific theoretical reference. Secondly, against LBD mechanism, we consider it as an important propagation mechanism, which can not only be incorporated in RBC model, but can be easily incorporated in DSGE models which are more complicated to expand the internal-propagation mechanism and the overall fit of DSGE models. As to maximum the method we provide, we can try to improve the LBD mechanism into more models related to financial institutions and financial market. For an instance, we can consider to incorporate LBD mechanism into monetary DSGE model, which improves the empirical result of monetary policy forecasting and makes a better practical guide.