Hello,

I have an issue with merging that I find rather cryptic.
I have several waves of household data that I merge by using the ID from the previous wave, which each household in the following wave is referenced to. But some households are not assigned an ID in the following wave, be it due to attrition. But Stata merges missing values and assigns households. As you can see below, several y3_hhid's are N/A (which I replaced missing values with), so they can actually not be assigned a y4_hhid as the merging ID is y3_hhid. Still, they have a value and this ID tends to be the same one. The shown y4_hhid for the missing y3_hhid continues for more than 200 observations. I really do not understand why Stata matches strings with different values.

Thank you for any suggestion!
Andreas

Array