2004 AHS Metro Survey

This document provides a brief introduction to the 2004 Metropolitan Survey dataset of the American Housing Survey.

• Click here to download the 2004 data in SAS Xport Format. The file is a self-extracting (*.exe, 12.9 MB) file, which expands to just over 98 meg. Updated on 11/02/05.

• Click here to download the 2004 metro data in ASCII format. The file is a self-extracting (*.exe, 23.5 MB) file that is a little over 23 meg in size, and expands to 8 files of size just over 400 meg. The data are in comma-delimited format. The first record in each file is a comma-delimited list of the variable names, in order. Updated on 11/02/05 .
• Click here to download the computer script (*.exe, 503 KB) that embodied the survey instrument (The Q Code).
• Click here to download Census Bureau's table specifications (*.exe, 79.5 KB) for its reports.
• 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. (Zipped MS-Word file, 167 KB.)
• Click here to download the SAS “file flattener” program which will convert the SAS data to a single file format in SAS. This program is customized for the 2004 Metro data. The size of the file is a little more than 7 KB.
• Comparer Files (*.zip, 478 KB): This zip archive contains documents that compare the 2004 AHS results to those of the previous survey for each metropolitan area. These files were produced as part of the quality control process and are provided to help users understand the changes in the AHS datasets. They should not be considered official publications.
• Descriptive Statistics (*.zip, 586 KB): This archive contains descriptive statistics by metro area (SMSA) for the 2004 American Housing Survey Metropolitan Sample (09/06/2005 release). These are intended to help analysts verify that they have set up the microdata files correctly. Statistics are provided for both the 8-file format distributed by HUD and the file produced by the "file flattener" utility. Descriptive statistics (minimum, mean, maximum, and sum) are provided in weighted and unweighed forms, by SMSA, for all numeric variables. Frequency distributions are provided for all character variables (not by SMSA). Updated on 11/02/05.
• Income Limits (updated on 5/3/07) (*.zip, 286 KB): This file contains estimates of Fair Market Rents (FMRs), income limits, area median incomes, and poverty levels for the records in the public use microdata of the American Housing Survey metropolitan sample for 2004. The data in this file can be matched with the AHS 2004 national file by using the CONTROL and SMSA variables.
• Topcodes File (*.xls, 225 KB): Certain variables are edited to protect the confidentiality of households with unusually high or low values. This MS-Excel spreadsheet lists the variables that are top- or bottomcoded. For each variable, the coding rule, maximum values, and minimum values are provided.

• Value Labels (*.zip, 35 KB): This zip archive contains code to add value labels to the character variables in the SAS version of the 2004 AHS dataset.
• AllTabs archive (*.zip, 708 KB): This zip archive contains text files of selected tables similar to the AHS printed reports. These were created by ICF Consulting as part of the quality control process. They are provided to give users a preliminary look at the data. They should not be considered official Census or HUD tabulations. There is one file for each of the metropolitan areas in the 2004 survey.

SURVEY DOCUMENTATION

Please download the Codebook for the American Housing Survey; Public Use File: 1997 and later (*.pdf).

Published Survey Results

Major Changes in the 2004 AHS

Income:

Whether or not a resident received SSI payments is no longer recorded in the variable QWELF (response category) along with public assistance or welfare payments. Now the receipt of SSI is tallied by itself in a new variable QSSI. (response category). The variable QWELF still exists but no longer includes SSI payments.

Alterations and Replacements:

Question HMR 7b was eliminated from the qcode. This question asked, “Was that wall-to-wall carpeting installed over existing, finished flooring, or was that put down over bare sub-flooring, such as concrete or unfinished wood?” It was asked if the respondent had indicated in HMR 7a that they had installed wall-to-wall carpeting (CARPET). In the past the combination of HMR 7a and HMR 7b produced the variables SUBFL (carpet installed), SUBFLWK (who did the work), and SUBFLAM (amount spent) for wall-to-wall carpet installed over sub-flooring. It also produced the variables FINFL (carpet installed), FINFLWK (who did the work), and FINFLAM (amount spent) for wall-to-wall carpet installed over existing, finished flooring. The variables SUBFL, SUBFLWK, SUBFLAM, FINFL, FINFLWK, and FINFLAM no longer exist. In 2004 the new variables CARPET (carpet installed), CARPTWK (who did the work), and CARPTAM (amount spend) exist for all units that installed carpet regardless of the type of flooring the carpet was installed over.