Proxy respondents between BHPS and UKHLS
Dear US help forum,
I merged the files of BHPS and UKHLS. When I did my computations, I noticed that there is an
amazing increase in the number of proxy respondents between UKHLS and BHPS.
In my sample, which contains singles who may or not transition to cohabitation or marriage, missing values
increase from 3,79% in the last wave of BHPS up to 9,85% in the first wave of UKHLS (and the trend in the following years is increasing).
Is there an explanation for this significant growth in the relative share of proxy respondents?
Thank you and best regards,