The table below is an experimental data product featuring monthly single-family housing starts for the West South-Central division. This innovative product is part of our effort to modernize economic measurement by leveraging new technologies. It demonstrates a new computer-vision approach to monthly economic measurement that can improve geographic granularity and accuracy while reducing burden on survey respondents.

For select areas of interest, Survey of Construction (SOC) data collected by field representatives have been replaced with estimates derived from satellite imagery and machine learning models to demonstrate the efficacy of this data collection method. This method utilizes high-resolution satellite images processed through Convolutional Neural Networks (CNNs) trained to identify construction stages like excavation, foundation, and framing. Please refer to the methodology document below for more details.

Currently, this experimental data product is on pause while work takes place on the development of improved modeling and methodology that will support increased accuracy and coverage of the product. No firm deadline has been determined as this process is quite fluid. The Census Bureau invites users to provide feedback on how to improve this experimental product to better meet user needs. Feedback can be shared by email at [email protected].