As President-elect Joe Biden’s transition team prepares to take office, it will be important to ensure all agencies have a strategy in place on how to effectively leverage data as a strategic asset. While the government established an overall federal data strategy in May, each agency still needs to identify specific steps for how it will connect its data, users and mission priorities. The Pentagon’s data strategy, released in September, offers a model for building...
As President-elect Joe Biden’s transition team prepares to take office, it will be important to ensure all agencies have a strategy in place on how to effectively leverage data as a strategic asset. While the government established an overall federal data strategy in May, each agency still needs to identify specific steps for how it will connect its data, users and mission priorities. The Pentagon’s data strategy, released in September, offers a model for building a data-centric organization with both a grounded foundation to support transformational change and a context-specific framework that brings the strategy to life. There are three key lessons Biden’s appointed agency heads can learn from the Defense Department’s efforts.
Encourage users to create innovative solutions by reusing existing datasets
Making datasets from different sources more useful to users is an effective mechanism for disseminating data to all the applications and services that can extract value from it. DoD’s data strategy seeks to move the department toward improved reuse of data by implementing globally unique identifiers that allows authorized users to easily discover, link, retrieve and reference data. Similarly, DoD is utilizing common metadata standards that allow data to be joined and integrated, further empowering authorized users to find innovative uses of existing data. While national open data policies have pushed agencies, such as the Department of Labor, to make significant progress in metadata management, agencies will struggle to scale change across their departments that have different maturity levels and data management implementations without consistent, standardized, practices and guidance.
Improve data accessibility by designating data stewards
When a dataset is not available or is out of date, it can be challenging for users to identify the correct agency contact to get in touch with. DoD is tackling this by assigning data stewards, data custodians and a set of functional data managers to achieve accountability throughout the entire data lifecycle. DoD’s data stewards are responsible for establishing policies governing data access, use, protection, quality and dissemination. Data custodians on the other hand, are responsible for promoting the value of data and enforcing policies. Finally, functional data managers implement the policies and manage day-to-day quality. While more than 80 other agencies currently have chief data officers, who are crucial to supervising the implementation of open data initiatives, many of these agencies still need to take the next step in designating stewardship over particular datasets and linking these identities with government personnel data.
Support AI and IoT strategies by creating capabilities to process and analyze new streams of data from emerging technologies
A data strategy is a precursor to an artificial intelligence or an IoT strategy. More importantly, the success of an agency’s data strategy will be directly related to how well it can leverage AI or IoT. DoD understands that fulfilling its ambitious AI plans is premised on users having access to high-quality training datasets, which is why DoD’s data strategy calls for the agency’s CDO, in partnership with department leads, to improve agencywide visibility of and access to these digital assets. To do this, the data strategy calls for the creation of a governance framework for managing the lifecycle of algorithmic models and associated data. The Department of Energy is still working on its own AI strategy, but without a similarly centralized effort to increase access to training data across all 17 DOE labs, the agency’s AI plans will likely struggle to create full impact. Similarly, while several agencies are exploring how IoT can provide solutions to their operational needs, few are well-equipped to process and analyze the new streams of data that would result from IoT adoption. Agencies will need to create their own frameworks outlining how to manage data responsibly so that federal IT leaders do not see collecting large amounts of data from the Internet of Things as a liability.
The U.S. federal data strategy has laid the groundwork for agencies to foster the use and power of data, but the Biden administration should use this as a springboard to articulate within each agency its own vision for a data strategy, tailored to its particular needs and challenges. Failure to do so may jeopardize the ability of federal agencies to keep up with new technologies.
Hodan Omaar (@hodanomaar) is policy analyst at the Center for Data Innovation, a think tank focused on the intersection of data, technology, and public policy.