Abstract: Open government data policies have become a significant part of innovation strategies in many countries, allowing access, use and re-use of government data to improve government transparency, foster civic engagement, and expand opportunities for the creation of new products and services. Rarely, however, do open data policies address intellectual property rights that may arise from free access to government data. Ownership of knowledge goods created from big data is governed by the default rules of intellectual property laws which typically vest ownership in the creator/inventor. By allowing, and in some cases actively encouraging, private capture of the downstream goods created as a result of open data policies, governments may fail to appropriate optimal returns to the public for its investment in big data. This Essay argues for coherence between open data policies and rules governing government ownership of intellectual property. It highlights the rule in US copyright law proscribing copyright in federal government works, arguing that public domain status is not invariably welfare-enhancing. The rule is sufficiently malleable to permit the federal government to assert ownership over knowledge assets developed from access to data that it owns or controls. Claiming copyright to engineer greater protection of the public interest could foster economic growth and facilitate the distributive welfare goals of intellectual property law more effectively than the public domain status that presumptively attaches to federal government works.