Testing Serial Correlation In Semi-parametric Additive Measurement Model And The Model’s Application In The Reaearch Of Residents’ Consumption

In recent years, the semi-parametric additive measurement error model is a kind of newly developmental semi-parametric statistical model, which takes the influence of measurement error into consideration. Including the advantages of linear and non-linear model, avoiding the “dimension curse” problem of nonparametric regression, this model has a strong modeling ability. It is a powerful tool for analysis of high-dimensional data. Therefore, in recent years, the research on semi-parametric additive measurement error model has become increasingly important for semi-parametric statistical model.So far, there are many experts and scholars have done some researches on estimating the semi-parametric additive measurement error model and its asymptotic properties. However, the testing for serial correlation of this model has not studied systematically. A good fitting effect model requires the fitted residual sequence to be the same independent variance. If the residuals are not independent, it indicates that the model has serial correlation. The existence of serial correlation will lead to a lot of statistical inference errors, such as the invalid of estimator and the failure of forecast. Therefore, this paper has presented the test statistics of serial correlation in the semi-parametric additive measurement error model. Numerical simulation indicates that the test statistics have a good test effect.Consumption is an important factor to promote national economic development. Since reform and opening up, China’s economy grew steadily at an average rate of 10%. However, in recent years, China’s economic growth mostly depends on investment and exports. A phenomenon of shortage in domestic demand present to the Chinese economy. In 2013, China’s final consumption rate fell to 49.8% and the final consumption boosted GDP growth by only 3.9%. The improving function of consumption has not been fully displayed. Apparently, simply relying on investment and exports will have a negative impact on China’s economic growth. Raising household consumption can not only maintain the sustainable economic development, but also improve the people’s living condition. Therefore, the residents’ consumption is important to the continuous growth of Chinese economy. Thus, we must vigorously promote domestic demand to promote the sustainable and healthy development of economy. Based on the provincial panel data from 2004-2012, this paper will use semi-parametric additive measurement error model to analysis the relationship between residents’ consumption and residents’ per capita disposable income, young and old dependency ratio and millions of per capita public transportation quantity.And serial correlation testing of the model shows it exists a significant serial correlation. After doing the researches, we get the following conclusions: firstly, raising the residents’ per capita disposable income helps to boost domestic demand growth in China. Secondly, raising young and old dependency ratio exacerbates the insufficient domestic demand in China. The problem of aging population in China needs to be solved immediately. This research will be of great importance in helping the government to formulate effective policies to stimulate consumption.