R help - Feb 2007 - Problems with obtaining t-tests of regression coefficients applying consistent standard errors after run 2SLS estimation. Clearer !!!!!

First I have to say I am sorry because I have not been so clear in my
previous e-mails. I will try to explain clearer what it is my problem.

I have the following model:



lnP=Sc+Ag+Ag2+Var+R+D



In this model the variable Sc is endogenous and the rest are all objective
exogenous variables. I verified that Sc is endogenous through a standard
Hausman test. To determine this I defined before a new instrumental
variable, I2. Also I detected through a Breusch Pagan Test a problem of
heteroskedasticity.

With the intention to avoid the problem of the endogenous variable and the
heteroskedasticity I want to apply first the technique 2SLS and then based
in these results I want to obtain the t-tests of the coefficients applying
Heteroskedasticity Consistent Standard Errors (HCSE) or Huber-White errors.

Like I showed above I have just one structural equation in the model. In
this situation, to apply 2SLS in R until I know there two possible ways:
First to use the function tsls() from package sem, or second, to use the
function systemfit() from package systemfit. I thought that systemfit  was
for situations when there are more than one structural equation in the
model. Anyway I probed with the two ways and I obtained similar results.
Below, I show the program lines:



*Program lines 1:*

* *

*>#### First option: applying tsls ####*

*>library (sem)*

*>Reg1 <-tsls (LnP~Sc+Ag+Ag2+Var+R+D,~I2+Ag+Ag2+Var+R+D)  *

*>summary (Reg1)*

* *

*>#### Second option: applying systemfit ####*

*>library (systemfit)*

*>RS <- LnP~Sc+Ag+Ag2+Var+R+D   # structural equation*

*>Inst <- ~I2+Ag+Ag2+Var+R+D    # instrumental variables*

*>labels <- list("RS")*

*>system <- list(RS)*

*>Reg2 <- systemfit("2SLS", system, labels, Inst,
saveMemory=TRUE) *

*>summary (Reg2)*



Now I want to obtain the t-tests of the coefficients but applying the HCSE.
I know two different ways to obtain them: First, applying the function
robcov() from package Design, or Second, applying coeftest() and vcovHV()
from packages lmtest and sandwich. The program lines are the following:



*Program lines 2:*



*>#### First option: using robcov() ####*

*>library (Design)*

*>options(scipen=20)*
*>robcov(Reg1) ### I tried with Reg 2 too*
**
**
**
***>####Second option: using coeftest and vcovHC ####*

*>library (lmtest)*

*>library (sandwich)*

*>coeftest (Reg1, vcov=vcovHC(Reg1 or Reg2, type="HC0"))  ### I
tried with
Reg 2 too*



In the two cases after trying to apply robcov or coeftest I obtained a
message of error:



*With robcov:*

*> Error in rep.default(1, p) : rep() incorrect type for second argument *

* *

*With coeftest:*

*>** **Error in terms.default(object) : no terms component*



In the following lines I copy all the sequence together from R. If I try
with systemfit instead of tsls I obtain the same result:



*Program lines 3:*

*>ibrary (sem)*

*> Reg1 <-tsls(LnP~Sc+Ag+Ag2+Var+R+D,~I2+Ag+Ag2+Var+R+D)*

*> library (Design)*

*Loading required package: Hmisc*

*Loading required package: chron*

*Hmisc library by Frank E Harrell Jr*

*Type library(help='Hmisc'), ?Overview, or ?Hmisc.Overview')*

*to see overall documentation.*

*NOTE:Hmisc no longer redefines [.factor to drop unused levels when*

*subsetting.  To get the old behavior of Hmisc type dropUnusedLevels().*

*Loading required package: survival*

*Loading required package: splines*

*Attaching package: 'survival'*

*        The following object(s) are masked from package:Hmisc :*

*         untangle.specials *

*Design library by Frank E Harrell Jr*

*Type library(help='Design'), ?Overview, or ?Design.Overview')*

*to see overall documentation.*

*Attaching package: 'Design'*

*        The following object(s) are masked from package:survival :*

*         cox.zph Surv survfit *

*        The following object(s) are masked from package:Hmisc :*

*         .noGenenerics .R. .SV4. *

*> options(scipen=20)*

*> robcov(Reg1)*

*Error in rep.default(1, p) : rep() incorrect type for second argument*

*> library (lmtest)*

*Loading required package: zoo*

*Attaching package: 'lmtest'*

*        The following object(s) are masked from package:Design :*

*         lrtest *

*> library (sandwich)*

*> coeftest (Reg1, vcov=vcovHC(Reg1, type="HC0"))*

*Error in terms.default(object) : no terms component*

*>*

I would like to solve these problems and to understand what the meanings of
the errors messages are. Also I want to know if there is another way to
obtain the HCSE through a function, because if I can not fix the problem my
last alternative is to do it manually.

I will thank a lot if somebody can help me

Best wishes

Guillermo

	[[alternative HTML version deleted]]