OSM Lab
Open Source Macroeconomics Laboratory (OSM Lab) Summer Boot Camp, Mentored Research, Open Source Economic Modeling, University of Chicago: Summer 2017 Boot Camp, Summer 2018 Boot Camp
I am the founder and Director of the Open Source Macroeconomics Laboratory (OSM Lab) at the University of Chicago, an innovative, intensive and immersive program that provides a seven-week computational economics boot camp to advanced undergraduate and graduate students from around the world. The OSM Lab also supports open source research and policy applications and funds mentored research assistantships for students. The OSM Lab has enjoyed generous financial support from a 5-year, $5.4 million grant from the Charles Koch Foundation, as well as institutional support from the University of Chicago's Becker Friedman Institute (BFI) and Social Sciences Division (SSD).
The OSM Lab has a five-point mission that involves supporting open source, open access teaching, research, and policy tools.
Create open access training material for computational economics
Support open source research that is collaborative, transparent, and replicable
Support policy-relevant open source applications
Support open source dynamic visualization tools
Support web apps for economic models
The central foundation of the OSM Lab is the seven-week boot camp that we hold every summer. In both years that we have held the boot camp at the University of Chicago, we have had around 150 applications from all over the world. From those applications, we fund and admit 25 students to participate as student researchers. OSM Lab pays these students a stipend of $4,200 and covers their travel and lodging during the camp. The OSM Lab also brings in top faculty in computational economics to provide open-access training in computational economic topics. The curriculum consists of a mathematics of optimization section, an economic theory section, and a computational and data science section. The computational section includes one of the best two-week, 16 lecture hour trainings on high performance computing (parallel programming) in existence. All curriculum materials for both the Summer 2017 and Summer 2018 boot camps are available in the OSM Lab GitHub repository.
OSM Lab has partnered with and I have joined the steering committee of QuantEcon.org, an organization that produces open access lectures and instructional material on macroeconomic theory and computational methods. QuantEcon was founded by Thomas Sargent (2011 Nobel Prize in Economics, New York University) and John Stachurski (Australia National University). Both Sargent and Stachurski have been involved in the OSM Lab at Chicago. A new platform that we have produced through QuantEcon is notes.quantecon.org. This site is a catalogue, search engine, and repository of Jupyter notebooks on computational topics. Anyone can submit a notebook. Notebooks are organized by topic and ranked according to views, comments, and up-votes. This allows a decentralized collection of training materials that get valued by the broad market of users. Curriculum from OSM Lab has been a part of this notes.quantecon collection from its beginning.
In addition to research, economic models that are used for policy analysis have a philosophical obligation to be open source. Closed source and proprietary models cannot be verified or tested for sensitivity or robustness. OSM Lab is currently supporting three models (OG-USA, Tax-Calculator, and B-Tax) that are being used for policy analysis (key examples of use are the U.S. tax reform of December 2017 and 2016 Presidential candidates).
Dynamic visualizations (Bokeh, D3, JavaScript) are a key tool for both advanced researchers as well as lay people to explore, digest, and understand the complicated results of heterogeneous agent models and other complicated models and data structures. OSM Lab is supporting active development of stand-alone visualizations collected in an online gallery, implemented as results in online web app tools (Cost of Capital Calculator), and a GitHub repository called plot-concepts where individuals can submit issues with dynamic visualization proposals and developers can help build those visualizations.
OSM Lab is partnering with developers to build front-end web applications that allow users to manipulate a key subset of model parameters, run its economic models on Amazon Web Services, and return results to the user via the web interface. Web applications allow researchers to do quick computations with economic models without accessing the source code. Web apps also allow non-economists (e.g., policymakers, journalists, businesses, lobbyists) to run an economic model without having to comprehend all the technical details. Our current web app portals include TaxBrain (microsimulation model for household taxation as well as three linked dynamic models including OG-USA overlapping generations model) and Cost of Capital Calculator which calculates the effects of corporate tax policy on marginal tax rates (incentives) for different industries and asset classes using the B-Tax open-source model. Finally, the OSM Lab at the University of Chicago is an outgrowth of a program that I ran for four years 2012-2016 at Brigham Young University--the BYU Macroeconomics and Computational Laboratory (MCL). The MCL had a similar curriculum and a similar mentored research mission as the OSM Lab.