1、如果是非平稳面板，但有不存在协整关系，是不是对变量差分后在回归？

2、面板的个体固定效应回归结果怎么会存在共同的截距项？？？（下图红色标示）用eviews6做的

Dependent Variable: LOG(RD?)    
Method: Pooled EGLS (Cross-section weights)    
Date: 08/08/08   Time: 13:36    
Sample (adjusted): 1992 2006    
Included observations: 15 after adjustments    
Cross-sections included: 28    
Total pool (balanced) observations: 420    
Linear estimation after one-step weighting matrix    
    
Variable Coefficient Std. Error t-Statistic Prob. 
    
C -1.918778 0.334629 -5.734048 0.0000
LOG(GDP?) 0.539796 0.053476 10.09413 0.0000
D? 0.292973 0.029934 9.787364 0.0000
LOG(TEC?) 0.503995 0.025046 20.12293 0.0000
LOG(RD?(-1)) 0.359358 0.033907 10.59849 0.0000
POP? 0.388713 0.428815 0.906483 0.3652
Fixed Effects (Cross)    
BJ--C 0.522298   
TJ--C 0.345165   
HEB--C -0.280948   
SX--C -0.042665   
NMG--C -0.163482   
LN--C -0.164931   
。。。。。