This grant funds the work of theoretical macroeconomist Xavier Gabaix, who is endeavoring to explain puzzling macroeconomic phenomena by importing into macroeconomic models insights gleaned from behavioral psychology. Contrary to the prevailing wisdom among macroeconomists, Gabaix’s work assumes human beings have limited attentional resources and must make choices about what to pay attention to and what to ignore. When attention is scarce, Gabaix argues, the pressing concerns of today crowd out consideration of distant tomorrows. This much microeconomists have known for some time. Gabaix’s contribution has been to show how this scarcity of attention and the consequent focus on the now can, in the aggregate, have predictable macroeconomic effects. Indeed, in early work Gabaix has used these assumptions to predict certain stubborn macroeconomic facts—like the absence of inflation in the U.S. despite years of low interest rates—that have vexed more traditional economic models. Funds from this grant provide three years of support to Gabaix to expand and continue this work. Supported activities include the testing and calibrating of Gabaix’s models against real-world data and the writing of a textbook that uses his framework to explain standard, well-understood macroeconomic phenomena.

This grant funds a project led by Alanna McCargo and Laurie Goodman at the Urban Institute’s Housing Finance Policy Center, to create a relational research database that links mortgage data made available through the Home Mortgage Disclosure Act with geographic and other data from the Census Bureau’s American Community Survey (ACS). The resulting dataset promises to provide economists and other researchers with a powerful new resource for investigating questions related to loan markets, geographic variations in housing prices, and consumer demand for credit. The Urban Institute team will design and implement a distributed, cloud-based architecture to house the database, and provide online computational access to the data through the Institute’s Spark Social Science computational platform. The team will also create and disseminate public guidelines and best practices for solving common problems with distributed, cloud-based data storage and the analysis of massive datasets.     In addition to the value of the dataset itself to researchers, the project will bolster the Urban Institute’s institutional expertise in addressing legal, security, privacy, and data acquisition and management issues related to large administrative datasets.