Exploring Housing Sales Data

by Bob Gradeck

September 18, 2019

We’ve been thinking about community indicators quite a bit lately thanks to several current projects and recent data requests we’ve received. As part of this work, we recently pulled property transactions for one and two unit properties (Note #1) shared by Allegheny County, and linked them with property assessment data to append property characteristics to each sales record. All of these datasets are available as open data on our Website. In this blog post, we share seven years of housing sales price data and also provide  a look at how ownership trends may be changing in different communities based on recent sales trends.

We hope these visualizations can  be used to better-help you interpret the data and understand how to design strategies in response to market conditions in your own neighborhood.

 

Median Sales Prices and Number of Sales

The first indicator we explored is median annual sales price for 1-2 unit properties. To calculate this figure, we did our best to include only arms-length sales, excluding “love and trust” and distressed transactions from our analysis. We were able to do this because Allegheny County classifies each type of transaction included in the sale description field. We developed a more-inclusive definition of valid market sales for this analysis (Note #2) than the one used by Allegheny County for the purposes of producing assessed values, as we noticed in looking at sales data that a large number of sales with prices comparable to valid transactions would have been excluded from our analysis had we relied solely on County sale classifications. 

You can see the median price and number of transactions data for City neighborhoods and County municipalities in our data visualizations between 2012 and 2018. 

Click on Each of the Following Housing Price Data Visualizations to Open them in a New Tab

          

City of Pittsburgh Neighborhood: Median Sales Price                      Allegheny County Municipality: Median Sales Price

 

           

City of Pittsburgh Neighborhood: Number of Sales                          Allegheny County Municipality: Number of Sales

In most communities in the City and County, home prices have increased, but the changes haven’t been consistent across all communities. Where prices have rapidly increased, communities have looked at adopting policies designed to prevent displacement. As researcher Alan Mallach wrote in Shelterforce, affordability challenges aren’t just tied to rising prices in communities like ours. Maintaining older homes can be expensive, and  incomes just haven’t kept-pace with inflation for many people. 

One key aspect of the housing market that we didn’t explore here has to do with rental prices. Unfortunately, there isn’t timely data available about how much people pay in rent in the U.S. Data from the American Community Survey is available, but at the neighborhood-level, data is grouped over a five year period, and the small sample size of the survey adds an element of uncertainty. While people have tried to use data from apartment listings as the backbone of rental data, this data is more-likely to capture information about  larger apartment buildings, not the smaller units typically found in many of our local communities. If you’d like to learn more about the shortcomings of rental data, check out a recent blog by the Urban Institute

 

Homeownership Status

We also wanted to see if we could learn how homeownership dynamics were changing in neighborhoods using the sales and assessment data. As a bit of an experiment, we developed an indicator we’re calling “Ownership Status.” Note – if you can come up with a better term, we’re all ears. To produce this indicator, we assembled data on all of the 1-2 unit properties that were sold at least once between January 2012 and December 2018.  For each of these properties, we looked at assessment data to see whether or not it was owner-occupied in October 2011 and March 2019. Unlike our analysis of median sales prices, we included all sales types, not just those that we classified as “valid”. We considered a property to be owner-occupied if it had either a homestead exemption, or if the property address matched the property owner’s address in the assessment data. Homestead exemptions provide a reduction in the County assessed value for owners that occupy a property as their primary residence. Since some homeowners don’t file for the tax exemption, we did not want to rely solely on the exemption as our measure of homeownership.

Our classification placed each property sold into one of four categories: 

  • Properties that were owner occupied in both 2011 and 2019; 
  • Properties that were not owner occupied in 2011, but became owner occupied by 2019,
  • Properties that were owner-occupied in 2011 but not in 2019, and 
  • Properties that were not owner occupied in both 2011 and 2019. 

In our data visualizations, we present the percentage of properties sold in each of these four categories for every City of Pittsburgh neighborhood and Allegheny County municipality. For context, we also included the homeownership rate across all types of occupied housing units as a label next to the community’s name in our data visualization. 

We also produced a second-set of charts showing the percentage of one and two unit properties that were sold to a new owner between January 1, 2012 and December 31, 2018. In most communities, between 30% to 40% of parcels were sold to new owners over this seven-year period.

This indicator makes it easier to see to the extent that homeowners or non-homeowners may be driving the market for one and two unit properties in individual communities. Community leaders in areas with considerable investor activity may consider policies and strategies to encourage investors to behave responsibly and maintain the condition of properties. These communities may also seek to stabilize, promote and incentivize homeownership. Community development stakeholders working in areas with strong homeownership trends may encourage developers to create more opportunities for renters to ensure that their communities might offer a more-inclusive housing mix.

Click on Each of the Following Ownership Status Data Visualizations to Open them in a New Tab

               

City of Pittsburgh Neighborhood: Ownership Status          Allegheny County Municipality: Ownership Status

                    

City of Pittsburgh Neighborhood: 2012-18 Turnover           Allegheny County Municipality: 2012-18 Turnover

Map of Parcel Ownership Status

More Data!

In addition to property sales and property assessment data, the Western Pennsylvania Regional Data Center’s open data portal also provides access to a number of datasets that can be used to better-understand neighborhood housing markets.  These datasets include: 

We have also built tools enabling people to access housing data more-easily, including our property dashboard, which links over six property data sources together, and the “Parcels N’At” tool which enables you to create a custom extract of parcel data.  We also provide technical assistance through office hours and via email, and we’d also like to hear your ideas for how we might be able to help you learn more about your community through data. 

Check-Out the Property Dashboard

Please let us know what you think.

 

Notes

  1. For all of the data appearing here, we limited our analysis to include only one and two unit residential properties based on their use code in March, 2019.  In the assessment data, the land use of the properties we included are coded as one-unit, two unit, rowhouse, and townhouse properties. We did not include condominium units, so the data presented here may not be as relevant or missing for markets like Downtown, Pennsbury Village, and the Strip District, where few if any properties are classified as one and two unit buildings. Since the data classifies “publicly owned” as a use code, we have excluded transfers of properties that were owned by a government agency, school district, or public authority in 2019. We also excluded properties with an assessed building value = 0 in 2019 to remove vacant property from the analysis in case there has been a lag in updating the use codes for properties that have recently been demolished.
  2. You can see the full list of what we considered to be a valid sale type, along with the types of codes excluded from this analysis. The property sales data dictionary can be found on our open data portal.

Tags: