The principal inputs to CreditForecast 4.0 are monthly aggregated credit files, organized by ZIP Code area; its output reaggregates the data to states and metropolitan areas.  We face two major complications:

• ZIP Code areas are subject to monthly change.  Maintained by the U.S. Postal Service for its own purposes, ZIP Codes designate delivery routes, which by their nature do not correspond in a simple way to any other geographic area.  Nonetheless, the USPS publishes a monthly file that maps from ZIP Code areas to counties, metro areas and states (see, for example, our February 2012 update).
• Metropolitan areas are subject to annual change.  Maintained by the U.S. Office of Management and Budget (OMB), "core-based statistical areas" consist of adjacent counties chosen by population and commuting patterns; metropolitan statistical areas (MSAs) and smaller metropolitan divisions are the most common types.  (See this summary of annual changes between 2006 and 2009 and this list of geo wild cards.)

CreditForecast 4.0 has a process to compensate for these changes:

1. For each dated credit report file, we use the corresponding dated ZIP Code mapping file.
2. We merge the dated mapping with a historical mapping, such that the most current values take precedence.  This helps account for possible address delays in a consumer's credit file.
3. We map the ZIP Code areas to counties (specifically, the FIPS codes for counties).
4. We aggregate county-level data to metro areas per the most recent OMB definition.