NLSY79 Appendix 6: Urban-Rural and SMSA-Central City Variables

National Longitudinal Survey of Youth - 1979 Cohort

NLSY79 Appendix 6: Urban-Rural and SMSA-Central City Variables

1979-1996 Urban-Rural Residence

Through 1996, the NLSY79 Urban-Rural Residence variables were constructed using the total and urban population data for the county of residence from the 1970 Census of Population Characteristics of the Population (for NLSY79 1979-1982) and from the 1980 Census of Population and Housing (for NLSY79 1983-1988). These data are included in the 1977 and 1983 County and City Data Book data files respectively.

The urban population consists of all inhabitants of urbanized areas. An urbanized area is defined as a central core or city and its adjacent, closely settled territory which have a combined total population of 50,000 or more (with exceptions in Alaska, New York, the New England states, and Wisconsin). These definitions have remained largely unchanged since 1950. For more detailed definitions and comments on exceptions, refer to the U.S. Bureau of the Census 1970/1980 Census of Population, Characteristics of the Population, Number of Inhabitants (Series PC80-1-A).

Calculation of the 1979-96 NLSY79 Urban-Rural Residence variables involved the following two steps:

  1. The percent of urban population was calculated by dividing the urban population of the county by the total county population and multiplying by 100.
  2. Rural counties for the NLSY79 variables were defined as those with between 0-49% urban population. Urban counties were defined as those with 50% or more urban population.

1998-2006 Urban-Rural Residence

Beginning with the 1998 release, the information needed to calculate whether the county was urban/rural was no longer available in the County and City Data Book releases. For 1998, a respondent is coded urban if living in an urbanized area or in a place with greater than 2500 population. There is no regard to county.

2008-2014 Urban-Rural Residence

For 2008-2014, the Urban/Rural variable is created using the definition of the US Census Bureau. A respondent is coded urban if living in an Urbanized Area or an Urban Cluster. All other respondents are considered rural. Respondents with quality of match of short street, long street or zip centroid are evaluated before assigning the urban/rural code. If the entire street or zip code falls within an area that fits the urban definition, the respondent is assigned an urban code. If the entire street or zip code falls within an area defined as rural, the respondent is assigned a rural code. If the street or zip code crosses an urban/rural boundary, the respondent is assigned an unknown code.

Non-interview respondents were assigned a valued of -5. Respondents who were residing outside of the United States were assigned a value of -4. Respondents who latitude and longitude of the current residence could not be determined were assigned a valued of -3.

User Note

The method of calculating urban vs. rural prior to the 1998 data release designated an entire county as either urban or rural. For example, if a county contained multiple respondents, some living in an urban-like area and others in rural areas, they all received the same designation depending on the degree of urbanization of the county of residence. In 1998-2006, the method of calculating urban/rural allows a respondent living in a rural area of a county to be coded rural, while another respondent in an urban area of that same county can be coded urban. The net effect appears to be an increase in respondents living in rural areas.

For all survey years, non-interview cases are set to a -5 value on the Urban-Rural Residence variable. All other missing cases are valid skips on these variables.

For a detailed discussion of the procedures used for hand-editing and merging respondent data with other data in creating the geographic variables, see Appendix 10: Geocode Documentation in the NLSY79 Geocode Codebook Supplement.

1979-2014 SMSA-Central City Variables

The NLSY79 SMSA-Central City variables are constructed using data for the SMSA/MSA and Place Description (PD) included in the City Reference File (CRF) data files. The 1973 CRF was used for NLSY79 1979-1982 SMSA-Central City variables; the 1982 CRF was used for 1983 variables; 1983 CRF was used for 1984-1987 variables; the 1987 CRF was used for the 1988-1992 variables; and the 1992 CRF was used for the 1993-1998 variables. See below for information on the 2000-2014 variables. (For a more detailed discussion of changes in official terminology and geographic designations and comparability across years, see Appendix 10: Geocode Documentation in the NLSY79 Geocode Codebook Supplement.)

1979-1998 SMSA-Central City Residence

Through 1998, calculation of the SMSA-Central City Residence variables involved the following two steps:

  1. A respondent's SMSA/MSA and PD were assigned based on the county, state, and zip code of current residence. (For a detailed discussion of the procedures used for hand-editing and merging respondent data with other data in creating the geographic variables, including the SMSA/MSA, see Appendix 10: Geocode Documentation in the NLSY79 Geocode Codebook Supplement.)
  2. Based on their PD and SMSA/MSAs, respondents were assigned to one of the following categories:
  • Respondents not residing in an SMSA/MSA were defined as "not in SMSA."
  • Respondents residing in an SMSA/MSA, but not in a central city of an SMSA/MSA according to the PD, were defined as "SMSA, not central city."
  • Respondents for whom the PD leads to an ambiguous central city residence status were defined as "SMSA, central city not known." These cases generally resulted from zip codes that cover more than one geographically-defined area.
  • Respondents residing in both an SMSA/MSA and the central city of an SMSA/MSA according to the PD were defined as "SMSA in central city."

Non-interview respondents were assigned a value of -5. Respondents with an ambiguous SMSA/MSA residence status or missing values for reasons other than non-interview are valid skips on these variables.

2000-2004 SMSA-Central City Residence

Beginning with the 2000 release, the calculation of the central city variable was revised slightly. The process still consists of two components. The first delineation is MSA/non-MSA residence. For respondents in an MSA, the residence is further defined by placement inside or outside a central city as defined by the Census Bureau.

Respondents who live in an MSA who have a quality of match code of either manual edit or zip centroid are evaluated before assigning the central city code. If the street (manual edit match) or the zip code area (zip centroid match) falls entirely within the boundaries of the central city they are coded as central city. If the street or zip code area fall completely outside the central city boundary, they are coded as not in a central city. Otherwise, these respondents are coded as central city unknown.

A further complication is that Maptitude, the software currently being used, does not include central city boundary files for several areas. Regardless of quality of match code, respondents in those areas are coded as being in a MSA with an unknown central city status since the mailing address is insufficient to reveal whether the respondent resides within the central city boundaries. When updates of Maptitude are available, the city boundaries will be updated.

2004-2014 CBSA-Principal City

The 2004-2014 CBSA-principal city variables were constructed using data for the CBSA (Core Based Statistical Area) and principal city. The creation of principal city variables involved the following two steps:

  1. A respondent's CBSA and principal city were assigned based on the latitude and longitude of the current residence (For more detailed discussion of the procedures, see Appendix 10: Geocode Documentation in the NLSY Geocode Codebook Supplement)
  2. Based on their CBSA and principal city residence, respondents were assigned to one of the following categories:
    • Respondents not residing in an CBSA were coded as "not in CBSA"
    • Respondents residing in a CBSA but not in a principal city of a CBSA were coded as "CBSA, not in principal city
    • Respondents residing in both a CBSA and a principal city were coded as "CBSA, in principal city"
    • Respondents with an ambiguous CBSA or principal city residence status were coded as"CBSA, central city unknown." These cases generally resulted from street or zip code that covers more than one geographically-defined area.

    Respondents who have a quality of match code of either manual edit or zip centroid are evaluated before assigning the CBSA-principal city code. If the street (manual edit match) or the zip code area (zip centroid match) falls entirely outside the boundaries of the CBSA, they are coded as"not in CBSA." If the street or zip code area falls entirely within the boundaries of the CBSA and the principal city they are coded as"CBSA, in a principal city." If the street or zip code area falls completely within the boundary of the CBSA but outside the principal city boundary, they are coded as "CBSA, not in a principal city." Otherwise, these respondents are coded as "CBSA, principal city unknown."

    Non-interviewed respondents were assigned a valued of -5. Respondents who were residing outside of the United States were assigned a value of -4. Respondents for whom latitude and longitude of the current residence could not be determined were assigned a value of -3.