Looking Beyond the Federal Data Strategy

There were quiet celebrations a year ago as the long-awaited Federal Data Strategy and a 2020 action plan came to fruition. It involved hundreds of dedicated people across the government and the action plan covered 20 specific elements. But the federal data strategy is but a foundation for a potentially broader vision.  A new report on the critical need for intergovernmental data sharing by Harvard’s Jane Wiseman says we need to extend it to become a national data strategy that embraces states and local governments as well.

Background. The Federal Data Strategy is an important framework that stitches together a number of different legislative and administrative initiatives into a coherent whole:

• the GPRA Modernization Act of 2010
• Open Data initiatives, such as data.gov
• the Digital Accountability and Transparency (DATA) Act
• the Grant Reporting Efficiency and Agreements Transparency (GREAT) Act
• the Foundations in Evidence-Based Policymaking Act, which includes the designation of Chief Data Officers and agency evaluation officers, and the creation of agency learning agendas.
• the efforts of external advocates for data sharing and usage such as the Data Coalition and Results for America

The Federal Data Strategy includes a ten-year roadmap for federal agencies and it is on the verge of releasing its 2021 action plan. While many of these federal-level initiatives rely on state and local data components, the current COVID pandemic demonstrates the need for a much more proactive intergovernmental data sharing strategy, per Wiseman.

Examples of Data Sharing Initiatives.  In her report, Wiseman showcases a number of successful initiatives that demonstrate the value of investing in data sharing efforts at the federal, state, and local levels, linking data from multiple sources across agency and jurisdictional boundaries:

• Federal:  U.S. State Department Links Data to Repatriate Americans Stranded Overseas During COVID-19 Crisis . In January 2020, the data team at the State Department began to bring together data from disparate public and private sources to create real-time information updates for department leadership on how to bring Americans home safely, first from Wuhan, China, and then from outposts around the globe. Under the leadership of Janice deGarmo, the department’s acting chief data officer, the data team quickly brought together all the data it could to help understand, monitor, and respond to the crisis, both from across the department and from external sources. Applying lessons from the Ebola outbreak, it synthesized data from CDC, Homeland Security, Customs and Border Protection, publicly available information, and their own on the ground intelligence to help repatriate both State Dept employees and others needing help. This effort led to the safe repatriation of over 100,000 Americans from 136 countries on over 1,100 flights working with embassies and consulates in every corner of the globe.
• State:  Virginia Leverages Its Data Trust to Rapidly Respond to COVID-19. Almost a decade ago, the state of Virginia began a state-local data sharing effort, starting with an opioid data project in one community.  Data were shared across state and local law enforcement, social services, judicial, and health agencies in order to better target prevention and treatment efforts.  In 2017, the opioid death rate finally began to decline in that community and effort was expanded statewide. This project demonstrated the value of broader data collection efforts and led in 2018 to the formalization of the role of a statewide chief data officer, the Commonwealth Data Trust, and the launch of the state’s Open Data Portal.  These foundational data sharing efforts paved the way for the state’s ability in early 2020 to quickly stand up a COVID-19 dashboard in a matter of days because it had already created a data sharing platform with data sharing agreements already in place with various state and local agencies in response to the earlier opioid crisis. This dashboard gives state leaders near real-time information about hospitals in need of supplies and locations with the largest COVID outbreaks.
• Local: Allegheny County, Penn., Uses Data to Prioritize Child Welfare, Homeless Services. Allegheny County, Pennsylvania (which includes the city of Pittsburgh) has built a data warehouse over the past two decades of social, health, justice, and education data that enables its case workers to prioritize the delivery of services where they are most needed.  For example, individual-level data has led to the development of risk models for the delivery of child welfare and homeless assistance services. For example, the Allegheny Family Screening Tool analyzed hundreds of data points from various data sources and is used by frontline case workers to predict the long-term likelihood of out-of-home foster child placements and whether to investigate a call about potential child endangerment.  It has also developed a similar predictive model to help prioritize the provision of homeless housing services, since the County only has resources to serve about half of its homeless population.  The tool uses existing client data from multiple sources to identify those individuals at the greatest risk of worsening mental health, incarceration, or emergency medical services.

Most Data Linking Strategies Also Leverage Digital Services. Wiseman found that the best government data sharing initiatives have a compelling vision of how data can be useful to transform operations – and are dove-tailed with the government’s digital service initiatives. She points to the Republic of Singapore as a leading example of how this is being done.  There, the government has created a seamless and customer-oriented government.  It provides a one-stop access for citizens to more than 300 digital government services from 110 different government agencies.

She observes that the role of the federal government is to start where it has – by defining a vision, principles, practices, and creating a governance framework (e.g., data sharing agreements, ethics, chief data officers), inventorying existing data, and setting standards. It has the ability to take on these tasks absent an urgent and compelling need.

However, state and local governments – because of limited resources -- tend undertake data sharing only when there is a compelling need to share data – emergency response, natural disasters, COVID response, floods, the opioid epidemic, etc.  However, they have learned that if you wait until you need the data, it is too late, because data collection, quality, and standardization efforts typically take years to cobble together. States and localities, however, are often more willing to support digital service investments that can show more immediate value to citizens. That is why linking data sharing to digital service transformation can help accelerate both.

Recommendations. Based on the findings from the relevant literature, expert interviews, and the case studies, Wiseman offers four recommendations to advance intergovernmental data sharing that extend beyond the existing Federal Data Strategy:  

• Congress and the president should create a policy and governance framework. They should define a broad data and digital excellence vision, with incentives to act, and a strong data governance structure that include states and localities. This would include actions such as establishing an “ask once” goal for data collection, rewarding agencies that link their data sets, and creating intergovernmental data councils.
• Congress and the president should establish funding and capacity building mechanisms to support implementation of increased data sharing across all levels of government. This would include actions such as supporting data literacy efforts in federal agencies and among federal leaders, funding for data sharing projects, and resources to improve data quality.
• The non-profit and philanthropic sectors should proactively support intergovernmental data sharing efforts. This would include actions such as providing incentives to innovate and link different sources or types of data between the federal, state and local levels, and supporting information exchange networks.
• Agency managers and data leaders at all levels of government should champion data sharing efforts. This would include actions such as articulating and creating a shared vision for data sharing, establishing shared data standards and protocols, and sponsoring communities of practice for data enthusiasts.

The COVID-19 crisis has demonstrated the urgency of investing in a national data collection effort. Performance data pioneer, Beth Blauer, and epidemiologist Jennifer Nuzzo recently noted in a New York Times op ed that eight months into the pandemic “there is still no federal standard to ensure testing results are being uniformly reported. Without uniform results, it is impossible to track cases accurately or respond effectively.” Without a national standard, states have developed their own approaches. Efforts like Johns Hopkins’ Coronavirus Resource Center try to standardize where possible, but overall, the national numbers are inconsistent. This hinders policymakers’ ability to smartly allocate resources, such as the vaccine, to where they will do the most good. Based on lessons from this effort, the incoming Biden Administration can leverage the lessons of the COVID-19 crisis to build on the foundation put in place by the architects of the Federal Data Strategy – and create a Golden Age of Data Sharing!