The housing data that powers our in depth housing authority and affordable rental property descriptions and cool data visualizations is a culmination of more than 10 years of work.
As as an affordable housing developer, our founder, David Layfield, often struggled with the housing data that was provided to the public by the Federal government. This data has for years been, and continues to be, delivered in several fragments through numerous databases published by multiple Federal departments.
More often than not, the same property will appear in more than one of these databases. Most of the time, there is no database key linking these properties together, creating many duplicates. Until recently, if someone asked how many properties in Texas have Section 8, Low Income Housing Tax Credits and USDA, Section 515 loan financing, the question couldn’t be answered.
For the past several years, we’ve worked with the Council for Affordable and Rural Housing (CARH) and the National Affordable Housing Management Association (NAHMA) on a housing advocacy tool that allows policy advocates to share affordable rental housing stats with their Congressperson in Washington. Until recently, we could only report that data in program silos due to the overwhelming quantity of affordable properties that cross programs.
When we learned that not only housing advocates like CARH and NAHMA using the housing data tool but Congresspersons and their staff were going to the web application seeking housing data for their own districts as well, we realized the power of good, reliable housing data. We vowed then that ApartmentSmart (Affordable Housing Online’s parent company) would be the one to improve the quality of the Nation’s affordable rental housing database.
In May of 2014, we rolled out the first version of this new database. In addition to multiple algorithmic, machine data processes, thousands of hours of manual deduplicating and fact checking have gone into the creation of this exhaustive database of America’s affordable rental housing.
Following are the data sources and a summary of the processes we went through to make the data as reliable as it is.
HUD publishes a database of LIHTC properties each year. This data is gathered through an exhaustive annual survey HUD conducts of housing finance agencies across the country. Mike Hollar, a Senior Economist with the Department of Housing and Urban Development manages the database and has done a great job of providing a picture of LIHTC properties.
This database is by far the most important one provided by the Federal government. It includes information not only about the LIHTC program but other funding programs that those LIHTC properties participate in. We rely on this database for stats on LIHTC, HOME, HOPE VI and CDBG.
We used machine geocoding and manual record matching to link properties in this database to records in other data sources. We will incorporate all updates to the LIHTC DB each year as they are released.
Our Project-Based Section 8 data comes from the Section 8 Contracts Database, part of HUD’s Multifamily Portfolio Datasets. The database is compiled by HUD on monthly basis from various sources in the Federal Housing Agency and its contractors.
We have linked this data source to all others using both machine and manual matching techniques. We update this dataset monthly shortly after HUD publishes its datasets online.
The one data source USDA, Rural Development does provide, which predates the Open Government Initiative, is the Multi-Family Housing Rentals web site on the USDA servers. The web site provides limited data on properties funded using Section 515 loans. The data provided is limited to location, age restrictions, total units, rental assistance units, bedroom mix and management agent.
The Section 515 Rural Rental Housing loan program is an important program in rural areas and should be included in any analysis of the country’s affordable housing portfolio. For that reason, we developed a proprietary web crawler that extracts the Section 515 property data from the MFHRentals site.
The data that is used for the MFHRentals site is linked to RD’s AMAS and MFIS data backends and is updated at least monthly.
We crawl the MFRentals site monthly to insure that our Section 515 data is as up to date as possible. In the future, we hope that RD publishes a more accessible and complete database that can be incorporated into our data tools.
Our Section 202 and 811 data is sourced both from the HUD Assisted and Section 8 Database and the LIHTC DB. The data is refreshed by HUD each month. We have utilized both machine and manual matching techniques to link this data to all other Federal housing data that is available.
The REAC score date we use for a property pages comes from HUD’s REAC dataset. HUD maintains two separate REAC score databases for public housing properties and other assisted properties like Section 8, Section 202 and Section 811 properties. The current REAC score dataset that HUD has provided is a little outdated with REAC inspections conducted through September 2009.
We will update our dataset when HUD releases an update to the REAC score data.
HUD does not provide a simple, complete source of data for its public housing properties that are owned and managed by public housing authorities throughout the country.
The data we provide for public housing properties was aggregated from the public housing physical inspection scores dataset and HUD’s Resident Characteristics Report (RCR).
We have used both machine and manual matching processes to link this data together.
We provide a great deal of data about our Nation’s public housing authorities including each housing authorities inventory of Section 8 and low rent units, contact info, PHAS score, SEMAP score, Section 8 waiting list status, copies of housing authority plans and resident characteristics data.
We provide up to date information on the status of hundreds of housing authority Section 8 waiting lists across the country. There is no central source of this data. We have utilized both machine and human search intelligence to assemble this one of a kind database of waiting list statuses.
Many of the features of our housing data tools include statistical data about the places where affordable rental housing is found. We rely on several government and private sources including the US Census Bureau, the United States Postal Service, Factual, Google Maps API, OpenCongress and others.
Some of the more specific data we provide on individual properties is data that we have collected for several years directly from affordable rental property owners and managers. This data has been aggregated by our staff and does not exist anywhere else. This data includes property photos, descriptions, office hours, unit square footages, rents, amenities and other similar data.
All of our data has been obtained and is utilized following the terms and conditions laid out by its owners. The vast majority of the data we utilize on our sites is in the public domain. In cases where there are specific license agreements, we make sure to follow the license terms or we don’t use the data.
We have worked very hard to merge into one distinct database, these countless sources outlined here. It has taken many people several years of hard work and smart computing to assemble this database.
ApartmentSmart.com, Inc. (Affordable Housing Online’s owner) has a copyright on this value-added housing database. Though databases generally are not copyrightable, “value-added” databases are. We have added a great deal of value to our database by undertaking multiple and exhaustive machine and manual validation processes as well as in the intuitive design of the data. Any use of the data contained on our pages without express written consent of ApartmentSmart.com, Inc. is prohibited.