By Eric Seymour
This second blog on our team’s research on corporate investor ownership of residential properties in New Jersey addresses the methods used and the techniques developed so far. While New Jersey provides a lot of property tax data for public and research use, combining records from different times and places into a single, usable dataset isn’t always easy. Hopefully, our experience can help others in the future.
The N.J. MOD IV Historical Database provides access to annual property tax data in New Jersey from 2010 to 2022. With a focus on neighborhoods with high or increasing levels of corporate-investor ownership, we needed a way to accurately locate both current and historical properties on a map. However, this process is trickier than it seems and can lead to significant mistakes if not done carefully.
Usually, this task involves matching tax data (like a spreadsheet) to a data layer that shows where a property is located on a map (using coordinates like longitude and latitude). But the ways in which properties appear in the tax data can change over time, which makes it difficult, or sometimes impossible, to match older properties with current map data based on common fields like block and lot numbers. Parcels can split, merge, or even disappear over time due to redevelopment or changes in local tax records. Additionally, block and lot numbers might be renumbered, making it harder to connect older data with current maps.
Through our work on seven New Jersey towns, funded by the Robert Schalkenbach Foundation, we developed a method to place historical properties correctly on the map — accounting for any renumbered blocks and lots. This process involved comparing older tax records with tables that contain coordinate information based on both current and past block and lot numbers, as well as property addresses. Simply matching old tax data with current block and lot numbers, which is common with newer data, can result in properties being misplaced, leading to wrong conclusions about neighborhoods with higher corporate ownership. Skipping the parcel data entirely and relying on street addresses can introduce new errors, especially if commercial mapping services have outdated or conflicting data (see our recent article for Cityscape, a journal published by the federal Department of Housing and Urban Development).
After spending time developing this mostly automated process, we now have the potential to expand our work on corporate investor ownership to the entire state. Essentially, our method lets us convert non-spatial tax data, like block and lot numbers and property addresses, into a spatial dataset that we can use to connect neighborhood-level corporate ownership with other information, such as housing market trends and demographic data from the Census and mortgage data from the Home Mortgage Disclosure Act (HMDA). In future blog posts, we’ll share our findings as we expand our efforts statewide.