to create lags with a panel dataset, do the following:
if "id" is your cross-section identifier & "year" is your time variable,
sort id year
tsset id year
* now you can create lags using the lag operator lN. for the Nth lag.
* to create the 5th lag of the variable x, then we do:
gen l5x = l5.x
* (that's the letter l and then the number 5) alternatively, fN. gives the Nth lead.
-- note, tsset only works if your cross-section identifier is a numeric
variable. if your cross-section variable is a string, then you'll have to
do something to turn it into a numeric variable first. one way to do this
is:
egen n_id = group(id)
take care,
chris
On Sat, 29 Mar 2003, Maria Popova wrote:
> Hi all,
>
> I have a time series/cross section data set and I'm trying to create 5-year
> lags for several variables. I have data on 158 countries between 1966 and
> 1997, but of course since not all countries existed for the entire period,
> different countries have data for different sets of years. I've sorted the
> data by time and within the time groups by country and then I tried
> creating the lags with:
>
> gen lag5_var=var[_n-790] (because 158*5=790)
>
> However, the resulting lag is incorrect... What am I doing wrong? Do I
> have to generate missing values for the country-years that are not in the
> data set? I thought Stata is supposed to do that for me...
>
> Thanks!
>
> *
> * 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/
>
*
* 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/
