The Theory, Method And Application Of Bootstrap Panel Unit Root Test

Panel unit root tests (PURTs) are a valuable tool for applied macro econometric research. And they are more powerful than traditional univariate unit root tests. Making use of the cross sectional dimension, however, raises specification issues regarding potential correlation among the cross sectional units. And the problem is focused on the access to the robust critical value of the test statistics under cross-section dependence. The national econometricians have studied that, the Bootstrap PURTs are different from but superior to the traditional methods based on the asymptomatic theory, which can obtain robuster critical value. In addition, the former have better small-sample properties and test power than the latter (Maddala and Wu,1999; Palm et al,2011). While under cross-section dependence, the problem is focused on the selection of different Bootstrap methods. Thus far, only few Bootstrap PURTs have been proposed in the literature. This paper is devoted to making a systematic research on theory, method and application of Bootstrap PURTs.Based on the theory of Bootstrap PURTs, this paper compares several Bootstrap methods via Monte Carlo test. And then, we apply Block Bootstrap PURT to test the hysteresis in unemployment in the labor market of our country, the result shows there exists hysteresis in unemployment in our country, which can provide beneficial guidance to manage the problem of unemployment.