Spectral density bandwith choice and prewightening in the estimation of heteroskadasticity and autocorrelation consistent covariance matrices in panel data models



dsso| wkh vdph rughu vhohfwlrq fulwhulrq wr erwk wkh YDU+p, prgho iru suhzklwhqlqj dqg
Wkh xqiyduidWh iDU+
pa)}‰ 1 PrghOv iru fkrrsiqj Wkh ])lι∣jin edqgzigWk ru Wkh fiqhvW vfdOhi
Wkhvh duh ghqrwhg dv SZ0TV+D, dqg SZ0TV+E, uhvshfwlyho|1

Lq rughu Wr vWxg| Wkh srWhqWldo khWhurjhqhlW| lq Wkh vhuldo fruuhodWlrq dfurvv Wkh furvv0
vhfWlrqdo xqlWv ri Wkh vhuldo fruuhodWlrq lq sdqho gdWd iru hdfk fdvh zh vWxg| Wkh iroorzlqj
yh vxefdvhv=

DUPD+7/7) DoWhuqdWlyhv=

' DUPD+7∕7)a g

DUPD+7/7)6 g

DUPD+7∕7)c g
<

DUPD+7∕7)d g

DUPD+7∕7)e g


Vit  '

—Pyit-n + %it + P%it-⅛>

l ‘> = ==>q.

Vit  '

Pyit-n + %n p%it-4>

l ‘> = ==>q>

Vit  '

— — PVit-n + %it + P%it-4,
[ %it>

l ‘ • > = ==>

l+ + I > =

n

2 >

==.q

Vit  '

P PvH-n + %it P%it-i,
[ %H>

l ‘• >===>ι

l+ +

>         2 l ɪ > = = = >

.

q.

Vit  '

— — Pyit-n + %it + P%it-4,
[ Pyit-
1 + %it P%it-4>

l ‘ • > = ==>

ln + i > =

n

2 >

==.q


+:19)


Zh hvWlpdWh e| Wkh zlWklq hvWlpdWru e= Zh h{dplqh Wkh ydulrxv hvWlpdWruv iru Wkh
dv|psWrWlf yduldqfh ri
e= Zh h{dplqh Wkhlu eldvhv/ yduldqfhv dqg PVHv1

Zh dovr h{dplqh Wkh vl}h ri Wkh w0WhvW ri

Kq g β ‘ f v.s.    Ka g β ' δ>

Wkh WhvW lv frqvWuxfWhg xvlqj Wkh zlWklq hvWlpdWru e dqg Wkh ydulrxv fryduldqfh hvWlpdWruv1

Wdeoh 4 uhsruWv Wkh eldv/ vduldqfh/ PVH/ dqg Wkh vl}h ri Wkh w WhvW xqghu DUPD +4/4)
iru
Eq> W ' E2D> 2 zlWk 'Ef=f> f=e> f=H= Wkh YDUKDF+D) kdv Wkh vpdoohvW grzqzdug
eldv/ Wkhq iroorzhg e| TV+E)/ TV+D)/ dqg QZ. Dv irxqg lq suhvlrxv vWxglhv lq Wlph vhulhv/
Wkh suhzklWhqlqj hvWlpdWruv ri SZ0QZ+D)/ SZ0QZ+E)/ SZ0TV+D)/ dqg SZ0TV+E) kdvh
odujhu PVHv Wkdq qrq0suhzlWkhqlqj phWkrgv. Lq jhqhudo/ TV phWkrgv kdvh vpdoohu eldvhv
Wkdq Wkhlu QZ frxqWhusduWv. Iru Wkh vl}h ri WhvWv/ YDUKDF+D) jdlqv d ehWWhu vl}h Wkdq uhvWv
ri phWkrgv exW kdv d odujhvW vduldqfh. Dovr/ TV phWkrgv kdvh jrrg vl}hv lq frqWudvW Wr QZ
frxqWhusduWv. Wkh suhzklWhqlqj phWkrgv/ krzhvhu/ kdvh uhodWlvho| zruvh vl}hv Wkdq qrq0
suhzklWhqlqj phWkrgv. Rq Wkh rWkhu kdqg/ eldvhv ri doo phWkrgv uhgxfh zkhq Wkh udqgrp
hhfW lqfuhdvhv. Wklv fdq eh revhuvhg WkdW PVHv iru doo phWkrgv Whqg Wr eh vpdoohu lq
fruuhvsrqghqfh Wr ghfuhdvh ri . Wkh vl}hv ri WhvWv vkrz qr vljqv ri lpsurvhphqWv hvhq
Wkrxjk PVHv ghfuhdvh. Wdeohv 5 dqg 6 vkduh Wkh vlplodu revhuvdWlrqv dv irxqg lq Wdeoh 4.
Wkh eldvhv dqg PVHv/ krzhvhu/ Whqg Wr ehfrph odujhu zkhq Wkh glphqvlrq ri frpelqdWlrq

43



More intriguing information

1. PROFITABILITY OF ALFALFA HAY STORAGE USING PROBABILITIES: AN EXTENSION APPROACH
2. Endogenous Heterogeneity in Strategic Models: Symmetry-breaking via Strategic Substitutes and Nonconcavities
3. Social Balance Theory
4. Economies of Size for Conventional Tillage and No-till Wheat Production
5. Do imputed education histories provide satisfactory results in fertility analysis in the Western German context?
6. Feature type effects in semantic memory: An event related potentials study
7. Micro-strategies of Contextualization Cross-national Transfer of Socially Responsible Investment
8. SOME ISSUES CONCERNING SPECIFICATION AND INTERPRETATION OF OUTDOOR RECREATION DEMAND MODELS
9. The name is absent
10. The name is absent
11. Sex differences in the structure and stability of children’s playground social networks and their overlap with friendship relations
12. The migration of unskilled youth: Is there any wage gain?
13. Emissions Trading, Electricity Industry Restructuring and Investment in Pollution Abatement
14. The name is absent
15. El Mercosur y la integración económica global
16. News Not Noise: Socially Aware Information Filtering
17. Effects of red light and loud noise on the rate at which monkeys sample the sensory environment
18. Wirtschaftslage und Reformprozesse in Estland, Lettland, und Litauen: Bericht 2001
19. The name is absent
20. The name is absent