Greater New Orleans Community Data Center
May 16, 2013
Trends in residential addresses actively receiving mail can serve as a useful indicator of the rate of change in occupied housing units. However, active residential addresses are not the same as occupied housing units. For the decennial census, the Census Bureau defines an occupied housing unit as the usual place of residence for an individual or group of individuals on Census Day (April 1). The U.S. Postal Service (USPS) defines a residential address as actively receiving mail if mail has been picked up within the previous 90 days. The USPS counts of active residential addresses may be higher than Census Bureau counts of occupied households in cities, and lower in rural areas. In cities, the USPS counts may be higher because they include each room in group quarters such as college dormitories, military quarters, and single room occupancies (residences for formerly homeless persons). The Census Bureau does not include group quarters in its count of occupied housing units. In rural areas, mail is frequently delivered to rural route boxes and P.O. boxes rather than street addresses, and thus may lead to an undercount compared to Census Bureau occupied housing units.
The Greater New Orleans Community Data Center (GNOCDC) relies on three different sources of U.S. Postal Service (USPS) data for indicators of occupancy and vacancy between the years of the decennial census. Each source includes different data formats but are derived from the same USPS Address Management Services database.
1. For ZIP code tabulations of addresses actively receiving mail, GNOCDC relies on the USPS Delivery Statistics Product. The Delivery Statistics Product is made available to the public by the U.S. Postal Service for a small processing fee, and includes ZIP code–level tabulations for vacant and possible residential addresses. The Delivery Statistics Product does not include tabulations of no–stat addresses. The U.S. Postal Service does not maintain an archive of its Delivery Statistics data‚ although some private resellers do.
2. For census block, census tract, and neighborhood–level tabulations of addresses receiving mail‚ GNOCDC relies on the Valassis Residential and Business Database‚ which must be purchased from Valassis Direct Mail Inc. Valassis is one of only a few companies nationwide that qualifies to receive weekly data feeds from the U.S. Postal Services’ Address Management Services database. The Valassis Database is unique from other sources of U.S. Postal Service data because it provides address level data. Although the address level data cannot be released to the public‚ GNOCDC developed a robust in–house geocoding process in order to generate extremely accurate census–block and neighborhood level data. The Valassis Database includes active and vacant addresses‚ but does not include no–stat addresses.
3. As a proxy for blight and unoccupied addresses‚ GNOCDC relies on HUD (Department of Housing and Urban Development) Aggregated USPS Administrative Data on Address Vacancies. HUD has an agreement with the U.S. Postal Service to receive quarterly aggregated data on the number of vacant and no–stat addresses. The HUD data is unique from other sources of U.S. Postal Service data because it includes tabulations of no–stat addresses‚ which can be used as a proxy for blight in some housing markets. HUD maintains an archive of its U.S. Postal Service data since late 2005, which is available to governmental entities and non-profit organizations registered as users. Data prior to 2008 aggregates residential and commercial addresses. In late 2010, the USPS deleted nearly every no-stat residential address from its New Orleans address list leaving the city with no indication of how many blighted homes remain.
There are some differences between the Valassis Database‚ HUD dataset‚ and USPS Delivery Statistics Product. The Valassis Database includes a record for each residence that is part of a “drop stop” (for example‚ each dorm room in a college dormitory). The Delivery Statistics Product and the HUD dataset‚ in contrast‚ count each drop stop as only one address regardless of the number of residences at the drop stop. This difference explains why the Valassis Database counts a few thousand more addresses than the HUD and Delivery Statistics Product datasets. There are also differences in reference periods between the sources.
You can find more detailed information in the following GNOCDC publications:
Valassis Lists Data as an Indicator of Population Recovery in the New Orleans Area. Ortiz and Plyer. 2008.
This technical documentation includes detailed descriptions of the Valassis, Inc. database‚ including a comparison of counts of active residential addresses between sources of U.S. Postal Service data.
Using U.S. Postal Service Delivery Statistics To Track the Repopulation of New Orleans and the Metropolitan Area. Plyer and Bonaguro. 2007.
This research note explains why U.S. Postal Service Delivery Statistics are useful for tracking repopulation in a post–disaster context‚ and documents some of the limitations of the data as a measure of repopulation.
Using administrative data to estimate population displacement and resettlement following a catastrophic U.S. disaster. Plyer‚ Bonaguro‚ and Hodges. 2009.
This paper published in an academic journal reviews literature from the fields of demography and other disciplines to identify available administrative data sets including USPS data that can form the basis of sound‚ relevant‚ and timely county–level population estimates following a catastrophic U.S. event.