Statalist The Stata Listserver


[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]

Re: st: RE: ivreg (2sls) to evaluate credit effect on farm production


From   george owuor <gowuor2001@yahoo.com>
To   statalist@hsphsun2.harvard.edu
Subject   Re: st: RE: ivreg (2sls) to evaluate credit effect on farm production
Date   Fri, 8 Dec 2006 14:16:58 -0800 (PST)

Thanks Austin i will try your suggestions.

thanks again

----- Original Message ----
From: Austin Nichols <austinnichols@gmail.com>
To: statalist@hsphsun2.harvard.edu
Sent: Friday, December 8, 2006 8:13:16 PM
Subject: Re: st: RE: ivreg (2sls) to evaluate credit effect on farm production


George--
First, get -ivreg2- via -ssc install ivreg2- and read the help file
and references.  On your specific problem, I cannot see why credit
would be included as an input, but I can see that it would determine
feasible levels of inputs such as fertiliser, hired labour, etc. and
why inputs and outputs would also affect the amount of credit
available, so that credit would not itself be a good choice for an
excluded instrumental variable ("instrument," commonly).  The first
thing you need to find is a valid instrument, i.e. something (or
preferably many things, so you can use -overid-) that affects
creditworthiness without being correlated with the error term in
lnY=f(inputs)+e (here e is "excess productivity," I guess).  Then you
have to worry about the weak instruments problem, in all likelihood
(see e.g. www.stata.com/meeting/5nasug/wiv.pdf).  You also probably
have some interesting dynamics, where running up debts impairs one's
ability to pay them off.  You might want to take a look at -xtabond2-
and the related paper explaining the method at
http://www.cgdev.org/content/publications/detail/11619. You might
benefit from thinking about a natural experiment where access to
credit is lumpy somehow (e.g. get credit or not, or get credit at
market rates or not) and the access is determined by things you can
see (components of a credit score of some kind), in which case maybe
you can use a regression discontinuity framework (see e.g.
http://www.urban.org/toolkit/data-methods/regression.cfm).

On 12/8/06, george owuor <gowuor2001@yahoo.com> wrote:
I have a problem. I am measuring credit amounts in a production
function, and given endogeneity of credit in that model i use ivreg,
but i am not sure if this is correct because ferterliser and  hired
labour are also in the same equation, yet credit affects production
through such inputs. Can ivreg sort out such a huge double endogeneity
problem. the equation is generally expressed as lnY= f(fertiliser,
hired labour, credit amount, other socio economc factors). could some
one assist to confirm if ivreg can solve such a problem.
*
*   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/

__________________________________________________
Do You Yahoo!?
Tired of spam?  Yahoo! Mail has the best spam protection around 
http://mail.yahoo.com 

*
*   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/



© Copyright 1996–2014 StataCorp LP   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index