As I had stated before, I wanted to update the Housing Profile Dashboard to offer a greater range of data, allow users to access the data, and improve the design to make the information more legible. I’m pleased to launch version 2.1 of the Housing Dashboard. Version 2.0 has been up for a few weeks, but after a few tweaks I am ready to give it a wider audience.
For those of you new to the Housing Profile Dashboard. Here is a quick rundown. This project seeks to provide a free tool for citizens, academics, and officials to collect the data for a Housing Needs Assessment. The dashboard uses data from the American Community Survey (ACS), Historical Census Data, HUD Picture of Subsidized Households, HUD LIHTC Property Level Data, and the HUD Consolidated Planning/CHAS Data to provide a picture of current housing need and demand. The Dashboard contains over 120 variables to help the user gauge the specific housing needs of their community. You can find more about the dashboard here. The dashboard, aside from major cosmetic changes, also adds additional features that aim to improve the ability of users to access and understand their communities.
Update 1: The Inclusion of Census Place Data
One of the limitations of version1 was that the location data was restricted to Counties. While some Counties have almost complete zoning control, most Counties take a backseat to incorporated cities. And even in Counties who retain zoning control, they often break their jurisdiction into specific areas. An example of this is Montgomery County, Maryland. While only select areas have their own zoning authority, the County uses local Sector/Area plans to guide most development. So areas such as Silver Spring, Bethesda, and Wheaton have their own Housing Needs. Similarly Cities such as Sacramento, California exist in larger Counties, making data at the county level unhelpful in gauging specific needs.
Census Place Data allows the user to look at data from the over 29,000 incorporated and census designated places in the United States. My hope is that by providing a more granular look at the data; local communities will have a way to gauge what the state of housing is in their backyard.
Update 2: Additional Detail in Housing Affordability
I’ve spent a lot of time thinking about how to gauge affordability. I’m still not completely satisfied with the method I have used but I have added some details which I think will improve the ability to get some insight into the hosing market. The first change is the inclusion of distributions of household income, owner costs, and rental prices.
This allows individuals quickly gauge the wealth of the population (is it mostly the wealthy? A mix?) and the relative costs of housing (for example, Newport Beach, California is exclusively high cost homes and a small sprinkling of above moderate income homes).
Another feature is the inclusion of a Net Gap and Gross Gap breakout for the Housing Gap Analysis. First, I should quickly explain the difference between the Net Gap and the Gross Gap. The Gross Gap is simply the number of housing units that have a price at or below 30% of the upper limit of a household in that income bracket. So if there are 1,000 households making 50% or less the median income. And there are 700 homes that have a rent that is 30% of the highest possible income in that bracket (50% of the median income), the gap is 300 homes.
However, one of the issues we see often is that there are households in higher income brackets that occupy lower cost housing units. The Net Gap adjusts for this by removing those houses occupied by higher income households. So let us assume of those 700 units, 200 units are occupied by those who make more than 50% of the median income. The Net Gap would subtract those units, making the gap 500 homes.
This difference allows communities to see how much of the housing market is being cannibalized by higher income households. One interesting takeaway is there are often surpluses of high cost owner homes, but often deficits of high cost rental homes. This can change what type of housing needs to be incentivized in a community.
Update 3: Downloadable Data
One item I also realize is that although I have tried to create a platform that allows people to easily see their information, it is not useable in other presentations or for pulling out specific variables. As a result, I have developed a backend process to allow users to download the data for themselves. Right now, you can download data for one place or county at a time. This is unlikely to change since the application is hosted using the free version of Heroku (and therefore subject to data limits), but individuals can now take the data I use and transform it for their own needs. The data downloads as a CSV file which you can load in Microsoft Excel or any other spreadsheet software.
Whats Next?
While I am happy with the current dashboard, there are still some updates I would like to pursue, utilizing additional datasets. In particular, I would like to try to highlight specific issues around housing that touch on specific types of housing. These include bringing in eviction filings, point in time homeless counts, housing quality, and creating comparison to the state and region.
If you are interested in helping with this project and have experience developing webapps in python, I encourage you to reach out. I have been working on the dashboard myself and while I enjoy working on it. I would like to push forward some other projects as well. Additionally, if you have any suggestions to improve the dashboard or additional variables to include, feel free to reach out an let me know.
Enjoy the Dashboard!