st: Clustering, Fixed Effects, or else?

[Leigh Lee <whaledandy@gmail.com>]

Hi,

I intend to estimate the following model:  Y(i,t)=X(i,t)+Z(t), where i
denotes units and t denotes months.  Y is updated monthly, and it
exhibits auto-correlation.  Some variables in the X matrix are updated
monthly, but some are updated quarterly.  Z is the variable I am
interested at, but it is time-variant only (by month).  What is the
best means to get reliable inference (standard errors) of the effect Z
has on Y.

To test my next hypothesis, I further want to estimate the following:
Y(j,i,t)=X(i,t)+Z(t), where i denotes units and t denotes months as
above and j denotes subjects in each unit i.  Once again, I am not
sure what is the best estimation method.

Thanks in advance for your help!!

Best,

Leigh
*
*   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/