American Community Survey (ACS) Employment
kawine opened this issue · comments
Predict whether an individual is employed, after filtering the ACS PUMS data sample to only include individuals between the ages of 16 and 90. See https://github.com/zykls/folktables. Copy the existing ACSEmployment setup in that repo, but add the RELSHIPP feature in addition to RELP, since that represents a coding change that was overlooked in the repo.
X: coded features (see dictionary here)
- 'AGEP': person age
- 'SCHL': schooling
- 'MAR': marital status
- RELP: relationship to head-of-household (old)
- RELSHIPP: relationship to head-of-household (new)
- 'DIS': disabled
- 'ESP': employment status of parents
- 'CIT': citizenship status
- 'MIG': mobility status
- 'MIL': military service
- 'ANC': ancestry records
- 'NATIVITY': nativity or foreign-born
- 'DEAR': hearing difficulty
- 'DEYE': vision difficulty
- 'DREM': cognitive difficulty
- SEX: sex
- RAC1P: race code
Y: ESR (employment status recode)
Domains:
- State
- Survey year (2014 – 2018)
What horizon? Person or household? https://github.com/zykls/folktables/blob/c4fc62bd357fcf19e2d5f4643f21aa931ef9e718/folktables/acs.py#L17
Also, should we be using race code as a domain / group? Survey year seems less interesting than race code (IMO)
cc @kawine