HMDA is a GET API that lives at
- You can query an entire dataset.
- You can query dataset slices, which are pre-loaded views we think are interesting.
- You can query concepts, which describe types of information found in the dataset.
The endpoint for querying all data begins with
|Endpoint||What it does|
||Gives all information about a dataset|
||Provides Home Mortgage Disclosure Act (HMDA) datasets starting from 2007|
Concepts are analogous to variables, or column headers in a spreadsheet. Concepts have properties, which describe all the possible values. You can also specify concepts in any of the supported file formats like so:
Think of slices as tables in a relational database. Every dataset has many slices representing different views of it, which you can use to construct advanced queries. The endpoint for
every slice is
You can even request a slice in HTML, XML, JSON, JSONP, or CSV. Just append the filename extension to the endpoint like so:
Putting it all together
Every year, the Federal Reserve finds interesting trends in HMDA data and publishes them in a report. Let’s say you want to replicate some of their 2012 highlights in JSON.
To compare refinances and home purchases in 2012, you would send the following query to the API:
To see changes in the FHA loan market for 2012, you would send the following query to the API: