Re: st: QUAIDS

[nigussie Tefera <nitefera@yahoo.com>]

Dear statlist users
Thanks for your all support, finally I could manage to modify poi’s nlsurquaids -- demand up date (2008)—to augment for zero consumption expenditure. In spite of econometric issues raised by Poi and other in our last e-mail conversation, the programme performs well on STATA 12 for up to five food groupings. However, it takes hours and hours if there are larger food groupings. I want to use for 10 food groupings, in my case, and couldn’t get the results yet (have been waiting for about 5 hours now). I wonder if there is ways (means) to speed up iteration a bit faster.  Any suggestion is highly appreciated.
Best
Nigussie

----- Original Message -----
From: Michael Musyoka <pmusyokastata@yahoo.com>
To: "statalist@hsphsun2.harvard.edu" <statalist@hsphsun2.harvard.edu>
Cc: 
Sent: Thursday, January 31, 2013 3:56 PM
Subject: Re: st: QUAIDS

Ben,
Yes, that is a key issue in demand system identification, that is, making the exclusion restriction.  The variables selected should follow the logical characteristics of an instrumental variables. The first stage-probit estimation of the system, is identified by inclusion of the equation's respective PDFs. Identification of the demand equations in the second stage is always a challenge as is with Heckman two Step Models. You would usually look for a variable that explains probability of 
purchases (usually referred to as market participation and computed as a
dummy variable from the budgets shares) but not demand in the second stage. More often than not, you would follow logic, literature and/or theory in identifying such a variable. Perhaps, you may want to have a look at the book by: Cameron and Trivedi 2005. Microeconometric Methods and Applications-Chapter 16 on Tobit and Selection Models, just to give you an idea.

Musyoka, 
pmusyokaus@yahoo.com 




----- Original Message -----
From: ben jones <jonesblw@yahoo.com>
To: "statalist@hsphsun2.harvard.edu" <statalist@hsphsun2.harvard.edu>
Cc: 
Sent: Thursday, January 31, 2013 12:34 PM
Subject: Re: st: QUAIDS

Dear All

Thanks for everyone's contributions, I have been following this exchange with interest. On a practical note might I ask those of you who have estimated this type of system, what sort of variables you have tended to use in the first stage probit which are not included in the conditional demand equation (I have struggled to successfully make this exclusion)?

Many thanks,

Ben
jonesblw@yahoo.com


________________________________
From: nigussie Tefera <nitefera@yahoo.com>
To: "statalist@hsphsun2.harvard.edu" <statalist@hsphsun2.harvard.edu> 
Sent: Thursday, 31 January 2013, 6:03
Subject: Re: st: QUAIDS

Thanks a lot Murat,

Nigussie 
----- Original Message -----
From: Michael Musyoka <pmusyokastata@yahoo.com>
To: "statalist@hsphsun2.harvard.edu" <statalist@hsphsun2.harvard.edu>
Cc: 
Sent: Thursday, January 31, 2013 8:45 AM
Subject: Re: st: QUAIDS

Murat,
Yes the budget shares are actually not scaled by use of the CDF....

More published literature include 

Two-Step Estimation of a Censored System of EquationsAuthor(s): J. Scott Shonkwiler and Steven T. Yen Reviewed work(s):Source: American Journal of Agricultural Economics, Vol. 81, No. 4 (Nov., 1999), pp. 972-982


Su, Shew-Jiuan B. and Yen, Steven T.(2000) 'A censored system of cigarette and alcohol consumption',
Applied Economics, 32: 6, 729 — 737...

Quasi- and Simulated-Likelihood Approaches to Censored Demand Systems: Food Consumption by Food Stamp Recipients in the United States
Author(s): Steven T. Yen, Biing-Hwan Lin, David M. Smallwood
Source: American Journal of Agricultural Economics, Vol. 85, No. 2 (May, 2003), pp. 458-478

Impacts of world prices transmission to domestic rice markets in rural Iran
M. Bakhshoodeh; Food Policy 35 (2010) 12–19




----- Original Message -----
From: nigussie Tefera <nitefera@yahoo.com>
To: "statalist@hsphsun2.harvard.edu" <statalist@hsphsun2.harvard.edu>
Cc: 
Sent: Tuesday, January 29, 2013 5:57 PM
Subject: Re: st: QUAIDS

Dear Murat,

Thanks a lot for your notice. I was completly wrong in saying so but I corrected that one in my latter e-mail.

Best

Nigussie

----- Original Message -----
From: Murat Genc <murat.genc@otago.ac.nz>
To: "'statalist@hsphsun2.harvard.edu'" <statalist@hsphsun2.harvard.edu>
Cc: 
Sent: Tuesday, January 29, 2013 3:12 PM
Subject: RE: st: QUAIDS

Dear Nigussie,

I have never seen a suggestion that left hand size variables (the budget share) should be multiplied by CDF.  

You may want to have a look at Lambert at al. (2006) "Fish and Meat Demand in Canada: Regional Differences and Weak Separability", Agribusiness Vol 22 (2) 175-199 for a good application of Quaids with Shonkwiler and Yen correction.  A more recent application is Ecker and Qaim (2011) "Analyzing nutritional impacts of Policies: An empirical study for Malawi," World Development Vol 39 (3) 412-428.

Murat

-----Original Message-----
From: owner-statalist@hsphsun2.harvard.edu [mailto:owner-statalist@hsphsun2.harvard.edu] On Behalf Of nigussie Tefera
Sent: Wednesday, 30 January 2013 12:33 a.m.
To: statalist@hsphsun2.harvard.edu
Subject: Re: st: QUAIDS

Dear Musyoka,

Both both the righ and left hand size variables should have to be multiply with CDF.  In fact, the PDF variables should also be added to both right and left hand side variables. In the left hand side we would expect equation like this "wi*CDFi + hiPFDi" instead of only "wi", where wi the budget share of commodity i. The CDFi and PDFi is expected to be derived from multivariate probit analysis following suggestion of Yen, Lin, and Smallwood (2003).

Because of such complexity, it is difficult for me to estimate Quaids agumented for zero consumption expeniture with CDF and PDF. I badly need suggestions from STATA userlist, partiuclary who have experienced and solved such problem. I am using Ethiopian Rural Household Survey (ERHS) panel data. Zero consumption expenditure is very server for the ten food groups I considerd for demand analysis. These include teff, barely, wheat, maize, sorghum, pluses, root crops, fruits & vegetables, animal products and other food. The first five items are the major cereals and staples in Ethiopian dish. 

Best

Nigussie 


----- Original Message -----
From: Michael Musyoka <pmusyokastata@yahoo.com>
To: "statalist@hsphsun2.harvard.edu" <statalist@hsphsun2.harvard.edu>
Cc: 
Sent: Tuesday, January 29, 2013 2:02 PM
Subject: Re: st: QUAIDS


Brian, 
I think you should check on whether it is ONLY the left hand side (dependent variables) which are the shares are multiplied with the CDF variable. I need to be corrected if my presumption that only the right hand side variables are multiplied by CDF is wrong.
Again, Brian, while I had you had exchanged some emails on nlsur QAIDs previously on the issue of zero budget shares, I had to do the zero budget shares correction manually as you suggest below. The results seem plausible though I did not drop one equation. 



Musyoka, 




________________________________
From: Brian P. Poi <bpoi@stata.com>
To: statalist@hsphsun2.harvard.edu 
Sent: Monday, January 28, 2013 5:25 PM
Subject: Re: st: QUAIDS

On 01/27/2013 05:48 AM, nigussie Tefera wrote:
> Dear Murat and Kolawole,
>
> Thanks a lot for your the help!
>
> As both of you rightly said, I am planning to multiply the CFD variables by all variables and it is only PDF variables that is going to be additional variables in the estimation techniques. The "nlsurquaidsNNP" program that I tried to write did exactly in the same way. The CDF is multiplied with the all variables and the PDF is the additional variables to the system. In my first post, I didn't spell it out properly, I am sorry for that. In fact, Shonkwiler and Yen (1999) warned that the share of expenditure will not be added up to one, as you also mentioned, and they suggest running full set of equations rather than for N-1 equations. However, Yen, Lin, and Smallwood (2003) as cited in Zheng and Henneberry (2010) suggested to recalculate share of one of the food items under study in such a way that the adding property to be hold. i.e. if  Si=wi+πPDFi . where si is the new budget share and wi the original budget share and PDF is the probability density
>   function derived from the first-stage multivariate probit function. Then they indicated the Sn=1- . I want to apply the same approach and exclude one of the food items in estimating the demand equations so as to control for singularity. And also as both of you mentioned correcting for the standard error is one of the difficulties in using this approaches. I think there are research works on progress to correct for it. In estimating the elasticities, again, we need to take into accounts the CDF and PDF variables and hence the standard formula for elasticity estimate doesn’t apply once the system is controlled for CDF and PDF variables.
>
> I prefer to use the recently released Boi’s quaids stata program to estimate the system (may be with modification) but I am not yet sure whether the program is already controlled for observed zero consumption expenditure as suggested by Shonkwiler and Yen(1999). Dear Dr. Poi, I need your suggestion in this regard.
>
> Best
> Nigussie
>

Nigussie,

Right now the -quaids- program recently published in the Stata Journal
(Poi, 2012) does not do anything with respect to the
zero-expenditure-share issue.

Moreover, -quaids- handles demographic variables using an extension of the
method introduced in Ray (1983).  My very strong suspicion is that
specifying a CDF, PDF, or inverse Mills ratio in the demographics() option
of -quaids- as if such a variable were just another demographic like the
number of children or the number of adults in a household is _not_ the
right thing to do.

I often get asked about Stata code for demand system estimation with
zero-expenditure shares.  I have nothing to show for that as of now.  As
you know, there have been many different methods proposed in the
literature.  I have never actually implemented any of them myself.

     -- Brian Poi
     -- bpoi@stata.com

References
==========
Poi, B. P. (2012). Easy demand-system estimation with quaids. Stata
Journal, 12, 433--446.

Ray, R. (1983). Measuring the costs of children: an alternative approach.
Journal of Public Economics, 22, 89--102.

*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/faqs/resources/statalist-faq/
*   http://www.ats.ucla.edu/stat/stata/

*
*  For searches and help try:
*  http://www.stata.com/help.cgi?search
*  http://www.stata.com/support/faqs/resources/statalist-faq/
*  http://www.ats.ucla.edu/stat/stata/


*
*  For searches and help try:
*  http://www.stata.com/help.cgi?search
*  http://www.stata.com/support/faqs/resources/statalist-faq/
*  http://www.ats.ucla.edu/stat/stata/

*
*  For searches and help try:
*  http://www.stata.com/help.cgi?search
*  http://www.stata.com/support/faqs/resources/statalist-faq/
*  http://www.ats.ucla.edu/stat/stata/


*
*   For searches and help try:
*  http://www.stata.com/help.cgi?search
*  http://www.stata.com/support/faqs/resources/statalist-faq/
*  http://www.ats.ucla.edu/stat/stata/


*
*  For searches and help try:
*  http://www.stata.com/help.cgi?search
*  http://www.stata.com/support/faqs/resources/statalist-faq/
*  http://www.ats.ucla.edu/stat/stata/


*
*   For searches and help try:
*  http://www.stata.com/help.cgi?search
*  http://www.stata.com/support/faqs/resources/statalist-faq/
*  http://www.ats.ucla.edu/stat/stata/  ;

*
*   For searches and help try:
*  http://www.stata.com/help.cgi?search
*  http://www.stata.com/support/faqs/resources/statalist-faq/
*  http://www.ats.ucla.edu/stat/stata/


*
*  For searches and help try:
*  http://www.stata.com/help.cgi?search
*  http://www.stata.com/support/faqs/resources/statalist-faq/
*  http://www.ats.ucla.edu/stat/stata/


*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/faqs/resources/statalist-faq/
*   http://www.ats.ucla.edu/stat/stata/