For directions on reading the variable description see Data Dictionary Introduction.

Availability: 
 

1850	1860	1870	1880	1900	1910	1920	1940	1950	1960	1970	1980	1990
X	X	X	X	X	X	X	X	X			X	X

Universe: 
1850 - 1950: All households and group quarters in identifiable cities. 
1980:            All households and group quarters. 
1990:            All households and group quarters in identifiable cities. 

Codes and Frequencies - Detailed

Description: 
CITY identifies the household's city of residence if the household resided in one of the cities identified in a given sample - see the comparability discussion below. To see which cities are identified for each year, check the codes and frequencies table below. The detailed code allows the identification of some cities that merged with others in the past, but most users will probably find this extra detail unnecessary. 

Comparability: 
This variable is essentially comparable for all years, but, as noted above, not all cities are identified in all years. A year-by-year discussion follows. All rank and size measurements refer to contemporary (not current) population figures unless otherwise noted: 

• 1850 and 1880: The city of residence is given if the household lived in one of the nation's 98 largest cities.
• 1860 and 1870: The city of residence is given for households in any city with 25,000+ inhabitants.
• 1900 and 1910: The city of residence is given for households living in any city with 25,000+ inhabitants. These years, however, identify in which population size range these smaller places belong and if they are located in New England (codes 8100-8810).
• 1920: The city of residence is given for households in any city with 25,000+ inhabitants.
• 1940 and 1950: The city of residence is given for households living in central cities of metropolitan areas only if the metropolitan central city and the remainder of the metropolitan area each had a population of 100,000+ (a total minimum of 200,000+) in 1980. In the 1940 and 1950 samples, cities can only be identified through central city status of a metropolitan area. Some metropolitan areas had more than one central city. So, in 1940 and 1950, the following cities share the same city code (i.e. they cannot be separately distinguished in the samples):

• 1980: The State, Metro, and Urban/Rural samples cover geography differently - see Chapter 2, "Sample Design" and the geography section of Chapter 1, "Introduction." The State and Metro samples each allow the city of residence to be given for households living in many, but not all, cities with populations of 100,000+. Note that the State and Metro lists of identifiable cities are not the same. In the Urban/Rural sample, however, only the central cities of urbanized areas are identified; see URBAREA for a further description of urbanized areas. In some instances in the Urban/Rural sample, urbanized areas have multiple central cities, and these cannot be identified separately. The following cities are combined for purposes of CITY in the 1980 Urban/Rural sample, and receive the code of the first city listed below (the largest):

• 1990: The State and Metro samples cover geography differently - see Chapter 2, "Sample Design" and the geography section of Chapter 1, "Introduction." Each sample divides the nation into Public Use Microdata Areas (PUMAs), containing between 100,000 and 199,999 residents. Note that the State boundaries are different than the Metro boundaries (see the IPUMS variable PUMA).

 

 
 
 
 
 
 
 
 
 

We constructed CITY from the household's PUMA of residence (contained in PUMA). While PUMA boundaries are sometimes coterminous with city boundaries, they are not specifically designed to be so, and they often are not - which means that some of the residents of a given city were grouped in a PUMA with residents of other contiguous civil divisions. Therefore, we usually identified the city of residence only if the PUMA(s) comprising that city were completely coterminous with the city. 

In some cases, we identified the city of residence even though the PUMA boundaries did not strictly follow city boundaries. These were PUMA(s) that we judged to be relatively "uncontaminated" with residents from outside the city. 

To be considered "uncontaminated," the PUMA(s) used in CITY had to: 

1. Contain within their boundaries 99-100% of all of the residents of the given city, and
2. Be 99-100% "pure," with fewer than 1% of the residents in the PUMA(s) living outside the city.
Therefore, when these PUMAs were translated into CITY, we are certain that they contain at least 99% of the city's residents and no more than 1% of the cases lived outside of the city limits. 

The following cities were identified in CITY even though the PUMA boundaries did not strictly follow city boundaries: 

• Sacramento, CA (State sample contains 99.99% of the city's residents);
• Akron, OH (State sample contains 99.6% of residents; Metro sample contains 99.47% of residents);
• Beaumont, TX (State sample contains 99.94% of residents);
• Atlanta, GA (Metro sample contains 99.99% of residents);
• Plano, TX (Metro sample contains 99.96% of residents).
• Phoenix, AZ (0.42% of Metro-sample residents are from outside the city; a smaller percentage of State-sample residents are from outside city);
• Huntington Beach, CA (0.70% of Metro- and State-sample residents are from outside city)
• Anaheim, CA (0.25% of State-sample residents are from outside city);
• Fullerton, CA (0.10% of Metro- and State-sample residents are from outside city);
• Hialeah, FL (0.07% of Metro- and State-sample residents are from outside city);
• Wichita, KS (0.23% of Metro-sample residents are from outside city);
• Pasadena, TX (0.02% of State-sample residents are from outside city).