Industry

Higher Education Research & Teaching

Language

Python

Features

Modeling
Numerical Computing
Subject Area Libraries

The aim of the Econ-ARK project is to make it easier to do ‘structural modeling’ of agents’ economic choices by providing a well-documented, open source codebase that implements the core techniques in a form designed to be adaptable to many different questions in both macro- and microeconomics. ‘Structural’ modeling of economic behavior aims to identify a rigorous mathematical (or computational) description of the rationale or decision rules that generate observed data, rather than just describing those data statistically. Econ-ARK aims to produce plug-and-play software modules that can be relatively easily combined, enhanced and adapted to address most economic choice problems.
The Econ-ARK originated as an initiative of the U.S. Consumer Financial Protection Bureau, with the aim of improving economists’ ability to model consumers’ financial decisions. It is being used for a number of ongoing research projects at CFPB. It also received early support and sponsorship from the International Monetary Fund, which is interested in improving its ability to model the consequences of heterogeneity and inequality across households, especially as they interact with public policy and social insurance programs. The Econ-ARK team has also given workshops at the Federal Reserve, at the CESifo conference, at the Computation in Economics and Finance conference, and at a joint conference sponsored by the US Treasury’s Office of FInancial Research, the Bank of England, and the Sloan Foundation; interactions with other central banks and policy institutions are ongoing, and researchers at several of these institutions are exploring the use of the toolkit for both policy and research purposes.

Be the First to Know

Be the First to Know

New developments and features from our sponsored projects, straight to your inbox, once a month.

New developments and features from our sponsored projects, straight to your inbox, once a month.