A Principal Components Approach to Cross-Section Dependence in Panels



are averages of RI disturbances, wt = i aiuit where uit = γηit + εit . Hence,
there is correlation between w
t and RII disturbances which also contain εit .
However, the weight of
εit in the former (ai) falls as N increases and so
does the endogeneity bias. Second, since
corr(xit, Xjt) = σ22 + σ2d), if we
let
σd2 0 then the regressors are identical across equations. In this case
plimT,N→∞ b = β + γδ and there are no gains from RII over RI. Third, al-
though the foregoing theoretical results (large T, large N consistency of b
) are
derived for a homogeneous panel with identical
β and γ across units, we con-
jecture that similar results apply to heterogenous panels. This is investigated
via Monte Carlo simulations in §4.

The variance of the POLS estimator of a =[bc]0 in RII is

var(a) E [(^ - α)(^ - a)0] = (X0X)-1E (X0vv0X) (X0X)-1      (15)

where X is an NT × 2 matrix. As shown above ρzw converges in probability
to 1 as T
→∞and N →∞. However, to the extent that wt is not a perfect
measure of the unobserved z
t in finite samples, there will be cross-equation
dependence in the disturbances, E
(vit,vjt) 6=0, and between disturbances
and regressors E
(vit,xjt) 6=0for i 6= j.HenceE(X0vv0X) 6= X0 E (vv0)X.
This suggests estimating
var(a) using an analogous formula to (13).

3.3 Serial dependence and heteroskedasticity

The assumption that dit and zt are iid processes is relaxed by letting

dit = Pddi,t-1 + εd,it, εd,it ~ iid(0, σd)
zt = pzzti + εz,t, εz,t ~ iid(0, σ2)

with -1 ρd, ρz 1. This introduces the serial correlation typical of eco-
nomic variables in the disturbances of RI (and RII). In particular, the matrix
E
(X0 vv0X) in (15) will have also nonzero elements E (xitvitxjsvjs) for t 6= s.
By letting
εd,it ~ iid(0, σdi) the heterogeneous variance of the idiosyncratic
in
fluence in xit introduces groupwise heteroskedasticity in RI and RII errors.
In particular, for RI we have

E(ui2t) = γ2(1 - δ)2σz2 + γ2δ2σ2d,i

By rewriting (6) as yit = βxit + γiwt + γi(zt - wt) + εit it follows that vit =
γ
i (zt - wt) + εit in RII. Hence to the extent that wt 6= zt , letting γi 6= γj will
also introduce heteroskedasticity in RII residuals.

11



More intriguing information

1. The name is absent
2. The name is absent
3. The name is absent
4. Can we design a market for competitive health insurance? CHERE Discussion Paper No 53
5. The Impact of Optimal Tariffs and Taxes on Agglomeration
6. Globalization and the benefits of trade
7. sycnoιogιcaι spaces
8. Income Mobility of Owners of Small Businesses when Boundaries between Occupations are Vague
9. The storage and use of newborn babies’ blood spot cards: a public consultation
10. News Not Noise: Socially Aware Information Filtering