On June 19, 23 smart young men and women—mostly undergraduates—will walk into a classroom at Saieh Hall in Chicago. Seven intense weeks later, they will emerge transformed into researchers capable of sophisticated computational economic analysis.
The students are taking part in the Becker Friedman Institute’s inaugural Open Source Macroeconomics Laboratory (OSM Lab) Boot Camp. They will spend up to 60 hours a week, counting lectures and assignments, in a collaborative crash course that blends computational skills, math, and economic theory, all focused on evaluating important policy questions.
“The idea is that after they do these seven weeks of training, they really come out as research monsters,” says OSM Lab director, Richard W. Evans. Participants acquire and apply the computational tools and specific knowledge needed to run and evaluate complex, general equilibrium macroeconomic models.
“We hit them with a firehose of information. They are putting in 60 hours a week, just finishing their assignments, but they come out with such a sense of empowerment and capability,” notes Evans, a senior lecturer in the Master of Arts Program in Computational Social Science at the University of Chicago and a fellow at the Becker Friedman Institute.
Evans observed such results running a similar program at Brigham Young University for four years. Part of his inspiration there was that it was difficult to find research assistants who could do what he needed to advance his own computationally intensive modeling. He found that after this boot camp, students were eminently hirable. What’s more, participants were so skilled their acceptance rates at top doctoral programs has been “phenomenal,” he adds.
Supported by a five-year grant from the Charles Koch Institute, the program provides students a $4,200 stipend and covers housing and travel costs. Additional funding will support ongoing research jobs for some participants. They will learn from and build connections with an expert faculty from Chicago and around the world that includes two Nobel laureates—the Becker Friedman Institute’s Lars Peter Hansen and Thomas J. Sargent of New York University.
Evans recreated the program at UChicago and fill a gap in economics education. Most PhD programs offer some computational training in commercial software packages like MatLab. “You don’t get broad training in programming in conjunction with the applications you need to solve economic models,” Evans says. Researchers “end up teaching ourselves how to code” by trial and error, focused on an immediate need or project, Evans explains.
In an era of big data and increasing computer power, “the modern economist is going to have to have computational skills, and it’s not going to be some proprietary software,” he says. “It’s important to have the breadth that comes with a true programming language like C or R or Fortran or Python.”
Encouraging Transparency
OSM Lab sessions will emphasize two themes: a focus on policy analysis and a reliance on open source tools and data that are available to all.
“In academic research, transparency is really important, especially if you’re doing computation. People need to make their code and data public,” Evans says. Writing code invariably involves mistakes; “that’s almost a maxim,” he adds. Freely sharing code means others can check it, run it, and confirm findings—which makes writing code that others can understand and replicate essential.
“When economic research informs policy, it’s even more important for computational research to be open, accessible, and tractable so that anyone can run the model, see the assumptions made, and see how sensitive results are to those assumptions,” he adds.Evans uses tax policy as a prime example; he has collaborated with the Open Source Policy Center at the American Enterprise Institute to combine their tax simulation tools with his general equilibrium models to assess economy-wide effects of tax policy. Such tools allow anyone—journalists, lobbyists, political candidates, or citizen taxpayers—to run scenarios on how proposed changes in tax brackets, rates, or deductions will affect government revenue and tax payments. These tools essentially crowd-source policy-making, giving anyone the power to test public statements and policy alternatives.
OSM Lab students will not be limited to evaluating tax policy; “nothing would make me happier than if they started populating different fields,” Evans says.
A Tough Program Students Love
Demand for the program was strong, drawing 150 applicants for about 20 funded positions. The first OSM Lab class at Chicago includes students from UChicago and Northwestern University Stanford and Berkeley, Brown and New York University, and even from China and Germany.
Testimonials from former BYU participants indicate that students find the program demanding but very rewarding. “Boot camp was a very stressful but also a very enriching experience. I really enjoyed being able to push myself beyond what I thought was possible,” wrote T.J. Canann, now a PhD student at the University of Minnesota.
Another key benefit many past participants valued was the necessity of partnering with their peers on assignments. “I really enjoyed the collaboration between students. The class was fast-paced and required that we developed a strong trust between classmates in order to work together and assist in each other’s learning process,” wrote Chase Coleman, who went on to the doctoral program at New York University Stern School of Business
Evans notes, “The intensive boot camp creates a situation where they absolutely have to collaborate to survive. I’ve never seen any of my students just be able to do it on their own. They come out of the program with a network of contacts and friends but also an understanding of how to work together in groups. That’s a key research skill, and, in this era of big, data-intensive projects with big research teams, it’s becoming more so.
The program is funded for five years, and with abundant computational hardware and expertise available on campus, Evans is already thinking of expanding the camp if he can without diluting the quality of the experience.
“I think we’ve found a niche that is not served, and it is just really powerful. With high-level, collaborative computational modeling, using full programming languages, and leveraging the hardware we have available, I think we are at the front of the wave that’s going to be important in the next 10 years.”
—Toni Shears