search across all the following databases:
Data and documentation
Common questions and answers.
Entire collection of data resources.
I have a student looking to measure the geographic dispersion of families (at least in the U.S., and other places if possible), specifically:
1) how many families with children at home have grandparents who are out of state and 2) how many divorced parents with children at home are not co-located in the same state (e.g., child lives with one parent and the other is out of state or child spends time with each parent in separate states)."
There seem to be lots of measures of an individual household's mobility (from the U.S. Census and CPS per se), and the U.S. Census even has a measure of grandparents living in the same household with children.
Do you know of any nationally representative data that will fit the bill?
There are 4 data sets that meet your needs: PSID, SIPP, NLS, and NSFH. These are all longitudinal data, but for any of these studies, a cross-section of a single year can provide you with a measure of geographic dispersion of families. Below are some details about each of the files:
The PSID has a Family Identification and Mapping System that allows one to link siblings together and/or parents and children. This would allow you to compare the location of grandchildren and grandparents or non-resident parents and children.
Here is some guidance on FIMS:
The PSID has a geocode file that would allow one to map the locations of these relationships. There are limitations on what one could show on a map for reasons of confidentiality, but some sort of tabular results by state/region would be quite informative:
x% of fathers live with biological child
x% of fathers live in same city/county as biological child
x% of fathers live in same state, but not same county as biological child
x% of fathers live in a different state
The Child Development Supplement has a "Fathers Outside the Home" file
http://simba.isr.umich.edu/VS/f.aspx [under CDS and TA (including Time Diary Aggregates]
Items included are:
Distance from child (in miles)
This would be helpful if the researcher wanted to get some quick results without going through the geocode application. However, this also limits the analysis to the CDS sample.
The 2004 SIPP panel has a topical module (Child Support), which has very detailed questions about child custody/child support arrangements:
Specifically, it has a question that asks where the other parent in the child support agreement lives:
Same county or city Same state (different county or city) Different state Other parent now deceased Other Unknown
There are also a fair number of questions about the non-custodial parent (on birth certificate or not; provides child support or not, etc.).
SIPP does not seem to have comprehensive information on non-resident grandparents, e.g., location.
It has a topical module on Child Care, which allows one to note how many grandparents are responsible for child care. However, it does not provide any information about the location of all potential grandparents - just by inference one is assuming that the grandparent providing care is nearby.
The NLSY97 has a question on which parent(s) Y lives with and distance from biological mother/father.
http://www.bls.gov/nls/handbook/2005/selvary97.pdf [Look under Youth History]
You can look at these items via NLS Web Investigator:
For instance, the "distance" question is in terms of miles not "different/same state." This is similar to the PSID's version of the same information.
In addition, as these youth age, they can become parents. There are some questions on the current residence of live biological and adopted children.
Starting with Wave II there are questions about the distance between children and the non-resident parent (in mileage). There is also a question about the mileage to the parents of the respondents (grandparents of the child). The distinction between paternal/maternal is not clear, but they do have an indicator on whether grandparents are alive.
Here is a link to the web version of the data, which allows one to explore the data:
The NSFH data include region as an item. The NSFH does have a geocode file, but its purpose is as a contextual file. The researcher never has access to the geographic indicators: