用levpet估计某一产业的TFP,命令为:levpet valueadded if industrycode==6 , free(l) proxy(m) capital(k) valueadded i(code) t( year) reps(50)
得到的结果如下:
Levinsohn-Petrin productivity estimator
Dependent variable represents value added. Number of obs = 40638
Group variable (i): code Number of groups = 594134
Time variable (t): year
Obs per group: min = 1
avg = 3.5
max = 10
v Coef. Std. Err. z P>z [95% Conf. Interval]
l .0720183 .0069313 10.39 0.000 .0584332 .0856034
k .171969 .009508 18.09 0.000 .1533337 .1906042
Wald test of constant returns to scale: Chi2 = 3512.37 (p = 0.0000).
红色字体处应该是有问题吧?number of groups的个数怎么与全部产业的企业组数相同,而不是只等于产业代码=6的?
并且用predict tfp, omega之后,发现tfp变量每个产业对应的值都被预测出来了,而不是我想要的产业代码6。说明这个命令是不正确的。
那问题出在哪里呢?