 Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at statalist.org.

# Antwort: st: Adding Up restrictions in LA/AIDS model

 From Richard Ochmann To statalist@hsphsun2.harvard.edu Subject Antwort: st: Adding Up restrictions in LA/AIDS model Date Fri, 19 Nov 2010 17:52:45 +0100

```...
well, in the AIDS, adding-up is fulfilled by definition, right? (see
Deaton, Muellbauer, 1980, An Almost Ideal Demand System, AER)
this is why - as you mention as well - you can only estimate J-1 equations
of a system of J equations, which you have done, as i undertand.
if this is the case, you cannot constrain explicitly on the Jth equation,
which you though try to do in constraint 1 to 7. how should stata find the
Jth equation? rather, adding-up constraints hold implicitly here when you
infer the results for the Jth equation residually. thus, you can just drop
constraint 1 to 7, if i understand your approach correctly.

best, rich

owner-statalist@hsphsun2.harvard.edu schrieb am 19.11.2010 16:57:59:

> Hello all,
>
> I have a problem about the correct way to make adding up restrictions
when
> I estimated the LA/AIDS model by SUREG. I estimated 5 commodities so I
> should run 4 equations.  The followings are my codes:
>
> constraint define 1 [q1]_cons+[q2]_cons+[q3]_cons+[q4]_cons+[q5]_cons=1
> //Alfa//
>
> constraint define 2 [q1]lnp1+[q2]lnp1+[q3]lnp1+[q4]lnp1+[q5]lnp1=0
>         //Gama//
> constraint define 3 [q1]lnp2+[q2]lnp2+[q3]lnp2+[q4]lnp2+[q5]lnp2=0
> constraint define 4 [q1]lnp3+[q2]lnp3+[q3]lnp3+[q4]lnp3+[q5]lnp3=0
> constraint define 5 [q1]lnp4+[q2]lnp4+[q3]lnp4+[q4]lnp4+[q5]lnp4=0
> constraint define 6 [q1]lnp5+[q2]lnp5+[q3]lnp5+[q4]lnp5+[q5]lnp5=0
>
> constraint define 7 [q1]lnmp+[q2]lnmp+[q3]lnmp+[q4]lnmp+[q5]lnmp=0
> //Beta//
>
> //homogeneity//
> constraint define 8 [q1]lnp1+[q1]lnp2+[q1]lnp3+[q1]lnp4+[q1]lnp5=0
> constraint define 9 [q2]lnp1+[q2]lnp2+[q2]lnp3+[q2]lnp4+[q2]lnp5=0
> constraint define 10 [q3]lnp1+[q3]lnp2+[q3]lnp3+[q3]lnp4+[q3]lnp5=0
> constraint define 11 [q4]lnp1+[q4]lnp2+[q4]lnp3+[q4]lnp4+[q4]lnp5=0
>
> //symmetry//
> constraint define 12 [q1]lnp2=[q2]lnp1
> constraint define 13 [q1]lnp3=[q3]lnp1
> constraint define 14 [q1]lnp4=[q4]lnp1
> constraint define 15 [q2]lnp3=[q3]lnp2
> constraint define 16 [q2]lnp4=[q4]lnp2
> constraint define 17 [q3]lnp4=[q4]lnp3
>
> global demand1 "(q1: w1 lnp1 lnp2 lnp3 lnp4 lnp5 lnmp)"
> global demand2 "(q2: w2 lnp1 lnp2 lnp3 lnp4 lnp5 lnmp)"
> global demand3 "(q3: w3 lnp1 lnp2 lnp3 lnp4 lnp5 lnmp)"
> global demand4 "(q4: w4 lnp1 lnp2 lnp3 lnp4 lnp5 lnmp)"
>
> sureg \$demand1 \$demand2 \$demand3 \$demand4, const(1 2 3 4 5 6 7 8 9 10 11
> 12 13 14 15 16 17) isure
>
> But STATA reports me that
>  (note: constraint number 1 caused error r(303))
> (note: constraint number 2 caused error r(303))
> (note: constraint number 3 caused error r(303))
> (note: constraint number 4 caused error r(303))
> (note: constraint number 5 caused error r(303))
> (note: constraint number 6 caused error r(303))
> (note: constraint number 7 caused error r(303))
> which means the code adding-up constraints have some problems .
>
> Can anybody point out why they are wrong?
>
>
> Thank you very much
>
>
> Xia
>
>
>
>
> *
> *   For searches and help try:
> *   http://www.stata.com/help.cgi?search
> *   http://www.stata.com/support/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/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/
```