分位数回归原理_面板分位数回归及Stata实现

1

简介

面板数据分位数回归所 使用到的命 令是 qregpd ， 由于这是一个 外部命令 ，因此首先需要先下载安装，我们来看一下下载安装出来的这样的一个结果

语法格式为：

 qregpd depvar indepvars [if] [in] [weight] , [quantile(#)               instruments(varlist) identifier(varlist) fix(varlist)               optimize(string) MCMC options Grid-search options ]

选项 含义为： depvar表示被解释变量 indepvars表示解释变量 [if]表示条件语句 [in]表现范围语句 [weight]表示权重 quantile(#)设置分位数，取值范围是0--1，默认为0.5，中位数。 identifier(varlist)设置个体变量名 fix(varlist)设置时点固定效应 optimize(string) 在估计qregpd时， Nelder-Mead (default)默认，自适应MCMC和 Grid-search 2

面板分位数回归案例

以 Stata 数据集为例来 进行面板数据分位数回归的案例操作。 该数据研究的是任职职位、工会组织对工资水平的影响。

该数据与Stata高级班中A10面板数据专题数据一致use "C:\Users\admin\Desktop\wage1810.dta", clearxtset idcode yeardescqregpd ln_wage tenure union, id(idcode) fix(year)

‍

. qregpd ln_wage tenure union, quantile(.1) id(idcode) fix(year)Nelder-Mead optimizationinitial:       f(p) =  -150.5734rescale:       f(p) = -.19788794Iteration 0:   f(p) = -.19788794  Iteration 1:   f(p) = -.19788794  Iteration 2:   f(p) = -.19788794  Iteration 3:   f(p) = -.19788794  Iteration 4:   f(p) = -.19788794  Iteration 5:   f(p) = -.19788794  Iteration 6:   f(p) = -.19788794  Iteration 7:   f(p) = -.19788794  Iteration 8:   f(p) = -.19788794  Iteration 9:   f(p) = -.19788794  Iteration 10:  f(p) = -.19788794  Iteration 11:  f(p) = -.19788794  Iteration 12:  f(p) = -.19788794  Iteration 13:  f(p) = -.19788794  Iteration 14:  f(p) = -.19788794  Iteration 15:  f(p) = -.19788794  Iteration 16:  f(p) = -.19788794  Iteration 17:  f(p) = -.19788794  Iteration 18:  f(p) = -.18066239  Iteration 19:  f(p) = -.15273439  Iteration 20:  f(p) = -.15273439  Iteration 21:  f(p) = -.10388595  Iteration 22:  f(p) = -.10388595  Iteration 23:  f(p) = -.10388595  Iteration 24:  f(p) = -.10388595  Quantile Regression for Panel Data (QRPD)     Number of obs:             19010     Number of groups:           4134     Min obs per group:             1     Max obs per group:            12------------------------------------------------------------------------------     ln_wage |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]-------------+----------------------------------------------------------------      tenure |   .0151846   .0024098     6.30   0.000     .0104614    .0199077       union |   .0824235   .0134725     6.12   0.000     .0560178    .1088292------------------------------------------------------------------------------No excluded instruments - standard QRPD estimation..