bNT = P1(Θ1,1 - Θι) (τmT Pt rm) Ui + Pι(el - e)ui
+ Pi(Θι,i - Θ1) TmmT Pt rm) Vi + Pi(ei — e)Vi.
From these three equations, it is obvious that the terms including Tm will be
the dominant factors determining the asymptotic properties of BNT, CNT, and
bNT. However, if the parameters Θ1,i are constant over different individuals so
that Θ1,i - Θ1 = 0, none of BNT, CNT, and bNT depends on Tm . For this case,
the asymptotic properties of the three terms depend on (ei — e). This result
indicates that the asymptotic distribution of the between estimator βbb , which is
a function of BNT, CNT, and bNT, will depend on whether the parameters Θ1,i
are cross-sectionally heteroskedastic or homoskedastic. Somewhat interestingly,
however, the distinction between these two cases becomes unimportant when
the model has no intercept term (ζ = 0) and is estimated with this restriction.
For such a case, Bnt, Cnt and bNτ depend on Xi instead of ei. With Xi, the
terms (Θι,i — Θι) and (e — e) in Bnt, Cnt, and bNτ are replaced by Θι,i
and ei, respectively. Then, it is clear that the trend term Tm remains as a
dominating factor whether or not the Θ1,i are heterogenous.
We now consider the asymptotic distributions of the within, between, GLS
estimators and the Hausman statistic under the two alternative assumptions
about the parameters Θ1,i.
CASE 1.1: Assume that the parameters Θ1,i are heterogeneous over differ-
ent individuals; that is, p1,2 - p12,1 6= 0. For this case, we can easily show:
plimN,τ →∞ T2m NTAnt = |
p1,2q1; |
1 1 plimN,τ →∞ T2m Nbnt = . ∙ 1 1 n plimN,τ →∞ Tm Ncnt = 1^ 1 и plimN →∞ n Hn = |
P1,2 — P1,1 (m +1)2 ; 0; σz2 , |
where qι = limτ→∞T P[(t∕T)m — T ∑t(t∕T)m]2 = R01[rm — 1∕(m + 1)]2dr.7
The first two equalities are obtained using the fact that limτ→∞ T Pt(t∕T)m
= R01 rm dr = 1∕(m + 1). In addition, we can also show that as (N,T →∞),
11
=⇒ N (0, p1,2q1σv2);
N Λ 2 P1,2 — P2
⇒ N I 0, σu / I -∣∖∙
u (m +1)
Tm √NTaNτ
ɪɪ b
Tm √N nt
7We can obtain these results using the fact that under given assumptions, √= Pt eit and
√= Pt Vit are i.i.d. over different i with N(0, σ2) and N(0, σ2), respectively, for any T.
More intriguing information
1. ARE VOLATILITY EXPECTATIONS CHARACTERIZED BY REGIME SHIFTS? EVIDENCE FROM IMPLIED VOLATILITY INDICES2. Database Search Strategies for Proteomic Data Sets Generated by Electron Capture Dissociation Mass Spectrometry
3. Life is an Adventure! An agent-based reconciliation of narrative and scientific worldviews
4. The name is absent
5. Literary criticism as such can perhaps be called the art of rereading.
6. Inhimillinen pääoma ja palkat Suomessa: Paluu perusmalliin
7. Reputations, Market Structure, and the Choice of Quality Assurance Systems in the Food Industry
8. ENERGY-RELATED INPUT DEMAND BY CROP PRODUCERS
9. The English Examining Boards: Their route from independence to government outsourcing agencies
10. I nnovative Surgical Technique in the Management of Vallecular Cyst