American Housing Survey (AHS)

Zones

Note on Documentation/Defining Zones
There is no documentation for how the zones were defined in the American Housing Survey (AHS). The idea of zones developed out of a request from the City of Chicago for submetropolitan data corresponding to their neighborhoods. After devising a custom set of zones for that metro area, the AHS team decided that they could do something similar for the other areas in the metropolitan survey. Zones have been defined by different analysts at different times, depending on when areas were first surveyed or resampled. Some of the guidelines used in these definitions include:

1. A minimum population size of 100,000 (this is a Census Bureau requirement).
2. Similarity in terms of structure type, tenure, income, housing cost, value, housing problems, and racial or ethnic mix.
3. Keeping central city and suburban tracts in different zones.
4. Keeping tracts in different states in different zones.
5. Keeping tracts in different counties (or New England townships) in different zones.
6. Spatial contiguity.

Given that these criteria usually can't all be satisfied, all but the first are violated to some extent. Different analysts have given different priorities to the criteria. Often, we have sought comments from local experts, such as HUD field economists, about where to draw the borders.

An additional problem is caused by the fact that metropolitan definitions change. When they do, sometimes areas are dropped from the survey while others are added. Because the AHS is longitudinal, we are limited in the way we can revise zones without violating confidentiality. Thus, the picture that we present now is based on a series of decisions by different people made at different times.