Merging Household Panel Data, but the roster order of individuals inside household change.
Hi
I try to measure intergenerational social mobility of individuals in a household. I use panel dataset of 1997 and 2017. The dataset is collected at a household-level, not individual level. I only have household id, not individual id.
The problem is when I merge dataset of 1997 and 2017 together, the order of a household roster change, due to some person move in or move out. This makes it impossible to track individuals overtime. From the householdid that I highlighted, 072136530905, the second person in 1997 becomes the first person in 2017.
Another problem is education attainment (hc7) changes as well, even it is not suppose to change because they are old people. This might come from an error when collecting data or data entry process. Gender also changes sometime. Age difference is not exactly 20, but 19-21.
Does anyone know how to solve this?
Thanks in advance
I interested in only parent in 1997, and children in 2017. There might be an easy way to to this. But if it possible, I would like to link all the individuals because I can use in other projects as well.
I try to measure intergenerational social mobility of individuals in a household. I use panel dataset of 1997 and 2017. The dataset is collected at a household-level, not individual level. I only have household id, not individual id.
The problem is when I merge dataset of 1997 and 2017 together, the order of a household roster change, due to some person move in or move out. This makes it impossible to track individuals overtime. From the householdid that I highlighted, 072136530905, the second person in 1997 becomes the first person in 2017.
Another problem is education attainment (hc7) changes as well, even it is not suppose to change because they are old people. This might come from an error when collecting data or data entry process. Gender also changes sometime. Age difference is not exactly 20, but 19-21.
Does anyone know how to solve this?
Thanks in advance
I interested in only parent in 1997, and children in 2017. There might be an easy way to to this. But if it possible, I would like to link all the individuals because I can use in other projects as well.
Comment