st: positive own-price effects

["Ivica Rubil" <irubil@eizg.hr>]

Dear all:

I have an issue not directly related to Stata, but presumably there are
a number of Statalist members who are familiar with the topic of my
interest. I'm estimating a complete demand system; QUAIDS, to be more
specific. The problem is that I systematically positive own-price
effects form most of the 13 goods in the model. I don't know what would
be the reason for such counterintuitive -- indeed, contra-theory --
results, but I suspect there is sth wrong with the prices that I'm
using, namely unit values (expenditure on good divided by physical
quantity of good) instead of genuine prices. Even when I tired, instead
of QUAIDS, a very simple specification like

regress w1 lnp1 lnp2 lnp3 lnp3 ... lnp13 lntotalexp

I get positive coefficient on lnp1. And similarly for all other goods.
Maybe it is due to the way in which I have assigned unit values to
households:

1. I have generated average unit values for year-region pairs for each
good (food items)
2. Since I need food items on higher level of aggregation (e.g., meat,
and not pork, beef, etc. separately) I have to aggregate somehow the
average unit values generated in step (1.). To do this I computed
weighted averages of these separate averages from the first step, using
as weights shares of each of the goods in the appropriate group of goods
(e.g., beef, pork,... in the group meat)
3. Such "prices" are then used instead of genuine prices.

Does anybody have any suggestion?

Thanks,
ir 

--
Ivica Rubil
Ekonomski institut / The Institute of Economics, Zagreb
a. Trg J. F. Kennedyja 7, 10 000 Zagreb, Croatia
t. +385-1-2362-269
f. +385-1-2335-165
e. irubil@eizg.hr
www.eizg.hr



*
*   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/