2003 AHS Data

Download the 2003 National AHS data SAS version (revised 02/14/06). (*.exe, 17.8 MB)

Download the 2003 National AHS data ASCII version (revised 02/14/06). (*.exe, 26.6 MB)

Q Code for the 2003 AHS (revised 5/21/04). (*.exe, 290 KB).
This file contains the computer script that constitutes the survey instrument.

Download the 2003 National AHS documentation (revised 02/28/06). (*.exe, 482 KB).
This file contains dataset content listings, descriptive statistics, frequency distributions, preliminary tabulations, comparisons with previous surveys, and the SAS "file flattener" program.

Documentation for HUD Income Limits (*.exe, 67 KB)
This file contains supplementary documentation for the estimates of Fair Market Rents FMRs, income limits, area median incomes, and poverty levels that are now included in public use microdata file (PUF) of the American Housing Survey national sample for 2003.

Table Specifications for the 2003 AHS (*.exe, 119 KB).
This zip archive contains a Microsoft Word file containing the Census Bureau's specifications for generating the tables used in the printed report.

AHS 2003 Value Labels File This file adds value labels to the character variables in the SAS version of the 2003 AHS dataset. (*.zip, 35 KB, Updated 07/21/04).

AHS 2003 Instrument Items File (*.exe, 204 KB)
The instrument items file reproduces all of the questions, prompts, and other information displayed during the course of an interview. Note that the fills, skip patterns, and branches built into the computer instrument make this document somewhat difficult to follow. However, the document does include all of the text used in the course of the interview.

Topcodes and bottom codes in the AHS public use file (*.xls, 29 KB)
In order to maintain confidentiality, unusually high (and, in a few cases, low) values in the AHS dataset are replaced with maximum (or minimum) values, called top (or bottom) codes. This MS-Excel spreadsheet documents these values for the affected variables.

2003 Worst Case Needs Tables (*.exe, 132 KB)
The Department of Housing and Urban Development, Office of Policy Development and Research (PD&R), publishes a series of reports on "worst case housing needs." The 2003 report was entitled "Affordable Housing Needs: A Report to Congress on the Significant Need for Housing." The primary data source for this report was the 2003 American Housing Survey. This archive contains the appendix tables of the report in Microsoft Excel spreadsheet format. PD&R provides these tables as a convenience to analysts who wish to use the tabulations in their own work.

SURVEY DOCUMENTATION

Major Changes in the 2003 AHS:

Race:

In 2003, multiple race classifications were introduced for which respondents, for the first time, could classify themselves in more than one race category, and the "other" category was eliminated in the edit of the answer category.

People were asked to respond to the question on race by indicating one or more of six race categories. The six race categories include:

• White
• Black or African American
• American Indian and Alaska Native
• Asian
• Native Hawaiian and Other Pacific Islander
• Some Other Race (this category is not read or displayed to the respondent)

Respondents who chose only one race are referred to as the race alone population. Respondents who chose more than one of the race categories are referred to as the Two or more race population. The 2003 edits eliminated the "other" race category. The 2003 American Housing Survey uses the complete Current Population Survey persons' edits. These edits do not allow "other" entries in race, but allocate one of the five specified response categories to those people reporting "other" race.

Previously many Hispanic householders (about 30 percent in 2001) stated that they were "other race," and 78 percent of the "other" race householders in 2001 were Hispanic. In 2003, although people may have reported themselves as "other" race, the edits allocated them to a different race category. So while over 7,000 people, 6,100 of whom were Hispanic, said they were "other" race, the edits assigned a category of "white only" to 92 percent of the Hispanics who had reported "other race." However, this ratio corresponds to what Hispanics say who do report a race category.

Weighting:

In 2003, the independent estimates (control totals) used to produce the weights are based on the Census 2000 with an estimate of change since then. This 2000-based weighting produces, on average, estimates that are about 1.0 percent lower then 1990-based weighting.

The 2003 AHS-N estimates are not available using 1990-based weighting. For comparative purposes 2001 data were produced using 2000-based weighting (the original data products used 1990-based weighting). As can be seen in the table below, the switch from 1990-based to 2000-based weighting produced a 1.0 percent lower estimate for 2001 at the United States level. The effect of the weighting change ranged from a 2 percent drop in the West to 2 tenths of a percent increase in the Northeast.

Detailed 2001 AHS-N data using 2000-based weighting are available from the Housing and Household Economic Statistics Division, U. S. Census Bureau, Washington, DC, 20233-8500 (301-763-3235).

Table: Total Housing Units in 2001 from the American Housing Survey Using 1990-based and 2000-based Weighting