# st: xtabond: Arellano & Bond 1991

 From "Gindo Tampubolon" <[email protected]> To [email protected] Subject st: xtabond: Arellano & Bond 1991 Date Tue, 2 Jul 2002 09:32:18 GMT1BST

```Dear All,

I am learning -xtabond- at the moment using David Drukker's last
year talk and the [R] Stata manual. I hope you can help me with
these questions:

1. I can't replicate the output of the Arellano & Bond 1991 Table 4.c.
( David and [R] reported the results in columns a1 and a2)
Where did I go wrong in specifying the model below as it gives
different estimates to A&B Table 4.c?
---
set matsize 1000
set mem 20m
use http://fmwww.bc.edu/RePEc/nasug2001/abdata.dta
tsset id year

forvalues yr = 1(1)9 {
local yr2 = 1975+`yr'
rename yr`yr' yr`yr2'
}

gen ln=l.n
gen l2n=l2.n

gen lw=l.w

gen lk=l.k
gen l2k=l2.k

gen lys=l.ys
gen l2ys=l2.ys

xtdes

xtabond n l(0/1).ys yr1980-yr1984, lags(2) twostep /*
*/    pre(w, lag(1,.)) pre(k)
--

Also, how do I get the estimates on Table 4.d?

2. A&B 1991 p.291 give the estimates of the long run
elasticity of wage and capital. My understanding is these
estimates are derived according to Wickens & Breusch
1988. My question is how to derive the standard errors? Or
is there an option to include to get these estimates and
their standard errors.

A&B 1991: Some tests of specification for panel data. Rev.
Ec. Studies, 58
W&B 1988: Dynamic specification, the long-run and the
estimation of transformed regression models. Ec. Journal,
98

FYI. Stata version 7 Special Edition On Windows NT 4

Thank you all very much for your help.
[email protected]
MIMAS
University of Manchester
*
*   For searches and help try:
*   http://www.stata.com/support/faqs/res/findit.html
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/
```