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

# st: survival analysis : computation of rate with spell data

 From pwetpublic@free.fr To statalist@hsphsun2.harvard.edu Subject st: survival analysis : computation of rate with spell data Date Sat, 13 Jul 2013 16:17:48 +0200 (CEST)

```Hello !

How I can compute the unemployment rate with spell data ?

I'm using Stata 12 for linux.

Here are my data (spells) :

id | Unemployment | begin | end  | duration (=end-begin+1)
1      0              0     15        16
1      1              16    80        65
1      0              81    132       52
2      1              0     25        26
2      0              26    132       107

The end time is 132 (11 years x 12 months).

What I want is something like that :
id | Unemployment | begin | end  | duration | Unemployment rate
1      0              0     15       16            0,5 (because between 0 and 15 the second individual is unemployed)
1      1              16    80       65            0,57  = [1*(25-16+1) + 0,5*(80-26+1)]/65
1      0              81    132      52            0 (because between 81 and 132 nobody is unemployed)
2      1              0     25       26            0,69 (= [50%*(15+1) + 100%*(25-16+1)]/26
2      0              26    132      107           0,25 (= [0,5*(80-26+1) + 0*(132-80+1)]/107)

The first solution I see is reshaping into time data (with one observation for each month), but I have too much variables, so that my computer cannot reshape.
And I don't see a second solution :p

Do you know how I can compute this ? Is there an function with stata ?

Thanks a lot !

Pierre
*
*   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/
```