In mySidewalk, IRS 990 nonprofit tax data is summarized by geography. This makes it easy for users to compare nonprofit activities in their communities, and to observe its interactions with data from other sources.
Nonprofit data are available at the
Tract,
Zip,
Unified School District,
State Senate District,
State House District,
Place,
County Subdivision,
County,
MPO,
Congressional District,
CBSA
State, and
Nation geographies.
There are some peculiarities in how this works that benefit from explanation.
How mySidewalk Builds IRS 990 Data
When mySidewalk retrieves data from the IRS, the EO BMF data table contains address data for every nonprofit organization in the nation.
How We Clean Address Information
Address data is self-reported by the nonprofits themselves, and is not always machine readable or easily translatable to a point location. To the extent possible, mySidewalk cleans the address data so that each address is formatted similarly (e.g. 123 Main Street, Townsville, MO 63038). This cleaned address data is geocoded, which involves assigning each address an XY-coordinate, which can then be mapped. There are limitations to the capabilities of geocoding software, namely:
PO Boxes, RR Boxes, and other similar non-address data cannot be geocoded. Organizations whose listed address is a PO Box or RR Box are only captured at the place, zip code, state, and nation geographic levels.
In some cases, PO Boxes belong to Zip Codes which represent a single post office location. Zip Codes are merely postal delivery routes, while their “shape” equivalents – Zip Code Tabulation Areas (ZCTAs) – may or may not correspond to these routes 1:1. Single-location Zip Codes do not typically correspond to a ZCTA. Learn more about this dynamic in our Geography Help Article.
To accurately locate nonprofits with Zip Codes that do not correspond to a ZCTA, we utilize a crosswalk table acquired from the Health Resources and Services Administration (HRSA).
Some addresses can only be geocoded to the accuracy of their home city. This data is only counted for the place, state, and nation geographies in mySidewalk.
Nonprofits with invalid addresses are only counted for the nation geography.
Given these limitations, counts between different geographies may not match. For example, adding the number of nonprofits in all counties in a state will likely produce a different total than the number reported at the state level.
Once nonprofit addresses have been geocoded, they are summarized for each geography by merely counting, summing, or averaging the data values that fall within each shape. I.e., to get values for the city of Alpharetta, Georgia, all nonprofit organizations whose geocoded XY-coordinates fall within the boundaries of that city are summarized.
What Geographic Data Means
mySidewalk’s 990 data summarizes nonprofit financial activity data by geography. It is important for users to understand that this data only summarizes data provided by organizations within each geography. This can be especially tricky when referencing revenue and expense data.
Revenue and expense data refers to the revenue and expenses of nonprofit organizations within a given geographic area. It explicitly does not refer to the origin or destination of those revenues and expenses. If we know that Nonprofit A raised $100 and spent $50 last tax year, that data is reported in the geography of Nonprofit A’s listed address. It is not known – and not reported or implied by mySidewalk’s data – where the $100 came from, or where the $50 was spent, geographically speaking.
Many nonprofit organizations work directly and exclusively in their communities.These communities may be constrained merely to their own neighborhoods, while others may be statewide or regional in nature. Some nonprofits have national scopes.
Given the nature of IRS 990 data, it is valid to say, for example, that “nonprofit organizations in my community had $x revenue and $y expenses.” By comparison, except in cases where local knowledge overrides this rule, it would not be accurate to say “nonprofit organizations in my community raised $x revenue from local sources,” or that “nonprofits in my community spent $y in my community.”