Recoding a variable from panel dataset with duplicate identifiers

[Romy Suter]

Hi everyone.

I am currently working on my thesis, covering gender and decision-making in South Africa's National Income Dynamics Study. I wanted some advice on how I can code the decision-making variable, in the face of duplicates.

In the survey, respondents were asked for the PID / identifier code of the person in the household who is the main decision-maker. I want to recode this variable to represent the gender of the decision-maker. However, when I run this code:

g decisiongender=0 if decisionmaker!=. & female==0 //male
replace decisiongender=1 if decisionmaker!=. & female==1 // female
label var decisiongender "decisionmakergender"
label val decisiongender decisiongender_dummy
label define decisiongender_dummy 0 "male" 1 "female"

The results represent the gender of the person responding, not the gender of the person that was nominated as the decision-maker.

I attempted a different solution by running the following:

preserve

keep pid female

rename (pid female) (decisionpid decisionfemale)

save "dec.dta", replace

restore

merge m:1 decisionpid using "dec.dta"
drop _merge
tab decisionfemale

However, I received the following error:

variable decisionpid does not uniquely identify observations in the using data
r(459);

I believe this is due to the fact of the data being a panel and having multiple responses for each PID over the multiple waves of the survey, as well as the survey allowing for a PID to be represented in multiple households. Dropping the duplicate PIDs will not aid my analysis.
Does anyone have any advice on how I could recode this variable? I am fairly new to Stata so there may be an obvious solution I do not know.
Thanks!

[Carlo Lazzaro]

Romy:
without an example/excerpt of your dataset, to be provided via -dataex- (as per FAQ), I find difficult to reply positively. Thanks.