Geocoding at the Center for Urban Research

What is Geocoding?

“Geocoding" transforms plain-old lists of street addresses into maps and mappable data that can be spatially displayed and analyzed. In simplest terms, geocoding uses a geographic information system (GIS) to assign geographic coordinates (such as latitude / longitude) to a list of addresses or other textual descriptions of specific locations. But once your list of (fill in the blank: members, students, survey respondents, program participants, etc) has been geocoded, then we can:

  • add value to your list by matching it with administrative and/or legislative districts - you can then sort and filter your list by district for more efficient outreach, mailings, policy updates, etc.;
  • map the concentrations of these locations - visualizing where to focus your outreach; and
  • analyze the proximity of your list to any other location that can be mapped: subway stations, Census tracts, health care facilities, etc.

CUR's experience and skill

The CUNY Mapping Service has extensive geocoding experience – collectively our staff has geocoded hundreds of thousands of addresses, mainly in New York City but also nationwide.

The latest improvement: Rooftop Geocoding

In 2009 we added “rooftop geocoding” to our capabilities – we can match your street addresses to actual building locations in New York City using the best publicly available data from city agencies. Most geocoding projects match your list of addresses to an approximate location along a “street centerline.” This is good if you need to match your locations with areas that use these “centerlines” for their boundaries (such as Census tracts or legislative districts). But tax parcel boundaries, for example, don’t extend to the center of a street – they typically stop at the curb or sidewalk. So street centerline geocoding doesn't help if you want to compare your locations with zoning, land use, ownership, building violations, permit information, etc. See the image below for a comparison

Rooftop Geocoding comparison

That’s where “rooftop” geocoding comes in. The city maintains a detailed database system with comprehensive address information for each tax parcel and building in the five boroughs. New York City’s addressing system can be complex – individual buildings often have multiple addresses, vanity addresses, names instead of street numbers, etc. -- and tax parcels often contain multiple buildings. CUR's staff has configured the city’s data so it works seamlessly with our geocoding system. You don't need to worry about importing the city's data and parsing it into a relational database structure - all you need to do is provide us with the addresses, and we provide you with tax parcel IDs, building identification numbers, and related information.

Contact us to discuss your project.

Preparing your data

Here are some things to consider:

  • We can match your addresses to different types of districts, such as City Council, State Legislature, Congress, as well as election districts, municipal court districts, Community Boards, etc. – especially in New York City but also statewide.

    • Addresses outside New York can also be matched, but we can discuss with you the types of districts we can match with.

  • Our geocoding system uses the most up-to-date and comprehensive street-level data from local planning agencies to ensure that geocoded addresses are matched with the correct districts (other geocoding systems can have a high rate of “false positives”).
  • You can provide your list in any electronic file format that uses rows and columns (such as Excel, Access, DBF, CSV, text delimited, etc). We will provide the list back to you in the same format, value-added with district IDs.
  • The only fields/columns that are necessary for you to include are street address, city/borough/county, and ZIP Code. If you need to link the district IDs for each address back to a larger database, include a unique ID number in the file. Don’t worry about removing apartment numbers, suite numbers, etc.
  • We typically match 85-90% of the records in your list, and can do better depending on time, resources, and address quality.

Contact us if you need help “cleaning” your data – identifying addresses that are hard to match such as PO Boxes, common misspellings, missing information such as street types or directional information, etc.