Large-N and Large-T Properties of Panel Data Estimators and the Hausman Test



/

(I1,1)

I1,21

I1,22

I1,31

I1,32

I1,33

0

/

I10 ,21

/ I20 1,21

I21,22

I21,31

I21,32

I21,33

I1,22

I0

I21,22

I22,22

I22,31

I22,32

I22,33

I0

I1,31

I0

I21,31

I0

I22,31

I31,31

I31,32

I31,32

0

I1,32

I0

I21,32

I0

I22,32

I0

I31,32

I32,32

I32,33

к

I10,33

I20 1,33

I0

I22,33

I0

I31,33

I0

I32,33

I33,33

0

0

0

where the partition is made conformable to the size of

00     0     0     0     0     0

x1,it, x21,it, x22,it, x31,it, x32,it, x33,it, zi .

We now consider each element in N Pi I1,i,NTI'1 i NT. For I1,1, notice that by

Lemma 15(a),

N X (xι,i - Exι,i) (xι,i Exι,i)0

i

N X IKi EXι,ik2 N X T X kxι,it - Exι,itk2 = Op (1).

Since each element in the diagonal matrix D1T tends to zero, we have

I1,1 = N X D1T (Xι,i Exι,i) (xι,i Ex1,i)0 D1T

= D1T Oi p(1)D1T=op(1).

Next we consider the second diagonal block of ɪ Pi Iι,i,NTI'ι i NT. Define

_ √τ ( X2,i Ex2,i ʌ

qiT = T     X3,i EFzix3,i J;

QiT = qiT qi0T .

Then, by Lemma 15 (b) and (d), for some constant q>1,

supE kQiT k2q = supE kqiT k4q ,

which verifies the condition (37) of Corollary 14. In consequence, from Corollary
14 and Assumption 8(i), we have

T ,X /(x2,i

N ÷ V'


x3,i


Ex2,i )(x2,i Ex2,i)0

EFzi X3,i) (x2,i Ex2,i)


(x2,i Ex2,i)(x3,i EFzi χ3,i)0

(x3,i EFzi x3,iXx^3,ii EFzi x3,i¢ '


= ⅛∑ τ∑∑
i ts


/ (x2,it Ex2,it)

× (x2,is Ex2,is)0

(x3,it EFzi x3,it)

× (x2,is Ex2,is)0


(X2,it EX2,it)        

× (x3,is   EFzi x3,is¢

x3,it EFzi x3,it    0

× (x3,is   EFzi x3,is¢ /


-   1           , 1∙   1 pn - Г γ22  Г23 A

= Ν∑Q∙T p lNm N XEQiT -(^ Γ23 Γ33 J,

ii

39



More intriguing information

1. Apprenticeships in the UK: from the industrial-relation via market-led and social inclusion models
2. Ex post analysis of the regional impacts of major infrastructure: the Channel Tunnel 10 years on.
3. The name is absent
4. DISCUSSION: ASSESSING STRUCTURAL CHANGE IN THE DEMAND FOR FOOD COMMODITIES
5. The name is absent
6. FISCAL CONSOLIDATION AND DECENTRALISATION: A TALE OF TWO TIERS
7. Input-Output Analysis, Linear Programming and Modified Multipliers
8. Endogenous Heterogeneity in Strategic Models: Symmetry-breaking via Strategic Substitutes and Nonconcavities
9. Artificial neural networks as models of stimulus control*
10. The name is absent