This is a periodic newsletter of the interesting things we’ve seen and what we are thinking about in open source policy analysis.

Upcoming Policy Simulation Library (PSL) DC Meeting. PSL DC is hosting its monthly meeting on Tuesday, January 29, at 12:00 p.m. at AEI headquarters in Washington, DC. You’ll hear updates from various members of the open source community and a presentation from Richard Evans (University of Chicago) on dynamic revenue analysis and using OG-USA, an open source, overlapping-generations model of the US economy. Link

The intersection of econometrics and machine learning. At the annual American Economic Association and American Finance Association joint luncheon, Susan Athey (Stanford) presented on the “The impact of machine learning on econometrics and economics.” Julian TszKin Chan – co-creator of the OSPC-incubated Policy Change Index for China and an attendee at Athey’s talk – told us, “Athey laid out an ambitious road map for applying machine learning and artificial intelligence in economic research.” Athey began by outlining the challenges faced when relying on machine learning (ML) algorithms for social science research. As ML inevitably permeates economic research, Athey foresees the application of econometric concepts to inform and interpret ML algorithms. To successfully employ ML to economics, Athey sees “economists as engineers.” As services, research, and education turn digital, it will be up to economists to understand context, define measures of success, and evaluate the algorithm. Link

Transparent tax policy model across the pond. EUROMOD is an open access tax-benefit microsimulation model of the European Union developed by the Institute for Social and Economic Research at the University of Essex. The model estimates the effect of taxes and benefits on incomes and work incentives for the EU as a whole and each member country in the EU, allowing for cross-country comparison. Last month, EUROMOD issued a new release that includes updated tax-benefit policies for all countries and new input data. Link

Open source analysis of federal minimum-wage policy. In 2014, the Congressional Budget Office (CBO) released a report on the estimated effects of raising the federal minimum wage on employment and family income. This past year, a team at the University of California, Berkeley, set out to recreate and open source the CBO’s research. Link

A model to make donations go further. Givewell, a charity evaluator, uses data to help maximize the impact of your charitable donations. The nonprofit has built, and open sourced, a model that evaluates the cost effectiveness of a host of charities. The model takes inputs such as administrative and program costs, academic research, and subjective and moral weights, and returns the estimated marginal expected value of the donation (e.g., cost per life saved). Adjust any of the model parameters, make structural changes, or use the default values to see where Givewell projects that donations will make the biggest difference. OSPC is not ranked by Givewell, and it is unclear how it might quantitatively value OSPC’s mission to further the adoption of open source policy analysis. Link and Link

Open data bill becomes law. In our last newsletter, we discussed the Foundations for Evidence-Based Policymaking Act. This week, that bill was signed into law by President Donald Trump. Link

Edited by Matt Jensen
American Enterprise Institute