Attaching individuals characteristics to household head

1

I am trying to attach the characteristics of individuals in a household to the household head. For example, if someone in the household is blind, I want a variable coded in the household heads line of data that says that there is an individual in the household that is blind.

The attach characteristics option in the extract only allows you to attach the household heads individual characteristics to other members of the house hold, not vice versa. And unlike the spouse option which codes both spouses with the others value, the household one only codes the household head to the individual.

I can do it in SPSS, but it doesn’t make any sense to me why it doesn’t create a variable that has individual characteristics for household members coded onto the household head in the attach characteristics when it does it for the spouse option. Granted, you could end up with a lot of variables depending on the family size, but seems like a pretty common issue. People want to look at households at a household level, but have information about the individuals in the household versus looking at individuals and wanting to know about the household head’s characteristics.

2

I understand your concern. It is relatively simple to attach the characteristics of a household head, since it is a unique individual for each household, or the spouse since there is only one spouse per person. But when you get into summary statistics such as “number of blind people in household”, especially when there is more than one category, the number of potential household-level variables becomes extremely large. Because of this, IPUMS does not attempt to anticipate every household-level variable that researchers might want to use. Instead, we provide all the household and person-level information needed to construct these variables yourself. We do provide some of the most commonly requested household-level variables which summarize person-level information. These include, for example, SSMC, NMOTHERS, and HHINCOME in IPUMS USA.

If you want every household member’s information on a single line, you can think of this as a “wide-format” dataset, with each variable repeated for each household member. You can reshape the “long-format” data available from IPUMS into wide format using most statistical packages.