Open-Source Policy Modelling

Computer models are widely used by government to analyze, predict, and evaluate the benefits, risks, and costs of existing and proposed public policy and regulations. The US Office of Management and Budget (OMB) has issed a series of guidelines for the conduct of such analyses. Among other things, they require that all US regulatory agencies ensure that their analyses are transparent and reproducible. This paper proposes extending these guidelines one step further: All computer models used by government to evaluate or justify public policy should be open source: That is, the source program code in which the models are written should be publicly available for anyone to download, review, run, and modify. This proposal is inspired by the remarkable success of open-source software — such as Linux, Apache, My SQL, and Firefox — and other open collaborative ventures, such as Wikipedia, the online encyclopedia. Open-source projects may have some interesting lessons for policy modeling: There are important similarities and differences between existing open-source software and model for policy and risk analysis. Possible benefits of open-source policy modeling include improved transparency, fewer errors, and greater reusability and extensibility. It could also lead to an expanded, more collaborative development process with a wider range of contributors, building more constructively on each others ’ contributions. A key issue will be the fate of current proprietary models. There will be benefits and challenges for each kind of organization affected: government agencies, stakeholder groups (including industry associations, environmental and social justice groups, and community groups), universities and nonprofit think-tanks, and consulting analysts and firms