[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

RE: st: movestay command

From   Zurab Sajaia <[email protected]>
To   statalist <[email protected]>
Subject   RE: st: movestay command
Date   Thu, 21 Jan 2010 11:15:12 -0500

Hi Sergio,
As it's outlined in the help for -movestay-, you can specify all three equations separately, e.g
movestay (y1 x11 x12 x13) (y2 x21 x22 x23 x24), select(s = z1 z2 z3 z4 z5)
in the original version of the program (SJ article) Xs from the main equations were automatically added to the selection equation, so z1-z5 (or just one z in that case) was interpreted as and extra instrument present in the selection equation (in addition tu all the Xs from the two wage equations) needed for identification.
But this was only in the original version of the program. I don't remember exactly since when(but for more then a year or so), the update of the program in SSC behaves slightly differently, allowing more flexibility:
now nothing is added automatically to the selection equation, you have to explicitly specify all the variables in there, so your Z vector should be the full vector of determinants of selection, not only instruments.
Hope this helps,

> Date: Thu, 21 Jan 2010 09:41:15 +0100
> From: [email protected]
> To: [email protected]
> Subject: st: movestay command
> Dear Statalisters,
> I'm using the movestay command (Lokshin and Sajaia, 2006) in order to estimate
> the effect of subsidies on
> some indicators of firms' performance.
> In the routine it must be especified the variables influencing in the
> selection equation (determinants of the subsidy) and
> variables influencing the firm's performance under two regimes: firms with and
> without subsidies.
> The question is, has the vectors of variables in the selection equation
> (probability to receive a subsidy) be the same of the vector of variables
> affecting firms' performance?. In theory there are important differences between
> these two vector of variables, but I can't found how can I
> use a different vectors of variables in this routine.
> Thanks for your help
> -------------------------------------------------
> This mail sent through IMP: 		 	   		  
*   For searches and help try:

© Copyright 1996–2024 StataCorp LLC   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index