Across nearly every public support mission, federal agencies are responsible for harnessing vast amounts of data. As government services are increasingly digitized, the massive amount of data retained by individual agencies can be overwhelming. However, when leveraged effectively, this data becomes a foundational component of effective, timely decision-making and improved customer experience, or CX, across the public sector.
To leverage data effectively and securely, the Federal Data Strategy directs agencies to “assess and proactively address the procedural, regulatory, legal and cultural barriers to sharing data within and across federal agencies and with external partners.” As such, the importance of data sharing and cross-agency collaboration took center stage at the Census Open Innovation Summit earlier this year.
The Census Bureau serves as the nation’s leading provider of data and insights about its people and economy, which means optimal data utilization and sharing is fundamental to its mission.
Like many data-rich or data-dependent agencies, the Census Bureau must balance maximum utility and security. In the case of the decennial census, the Bureau is expected to share the data it collects to improve a myriad of federal programs, employee experience, and overall CX. These objectives are critical but cannot come at the expense of privacy and confidentiality.
The key is determining which barriers to data-sharing are redundant or unnecessary and which are essential to preserve the security of sensitive personal and government information. To strike this balance, agencies such as the Census Bureau will need to engage in continuous, extensive discussions with one another and with industry partners.
Means to an end
As a starting point, agencies should look at data as a means to an end. Data users must assess their final desired outcomes and determine the minimally invasive data elements that meet them.
Too often, data recipients, especially in cross-agency data-sharing relationships, want the flexibility of receiving raw data to process and utilize to meet their changing needs. While this is an understandable desire, it must be balanced against the paramount need for personal data security. Starting with the desired outcome is an effective approach to resolving this conflict; asking data providers to contribute insights and outcomes rather than raw data makes it far easier to resolve data privacy concerns.
For example, while the Census rightfully protects individual data, aggregated data can often be just as useful while preserving privacy. It is challenging to decide what data best meets desired outcomes and what degree of risk is acceptable to meet them. Fortunately, differential privacy and other approaches can provide structured ways for determining how much and what type of aggregated data or insights can be shared while mitigating individual privacy risks.
Data governance around sensitive and non-sensitive data can significantly reduce data-sharing barriers. Agencies that distinguish appropriately between sensitive and non-sensitive data and label accordingly foster interoperability and help agencies make decisions about their data faster. By working with trusted industry partners to create a data governance strategy, agencies can be more deliberate about data integration and security decisions, breaking down inter-agency and intra-agency siloes.
In addition to privacy concerns, cybersecurity risks must be considered. The federal government is subject to a barrage of cyberattacks daily. As such, agencies adhere to stringent cybersecurity protocols, which can restrict data access. All federal data architectures should be shrouded in cybersecurity measures that account for requirements such as zero trust and the Evidence Act.
Fortunately, there are innovative solutions to keep data secure yet accessible. Data fabrics or data meshes allow agencies to leave data where it is and expose it through APIs. This allows the information to remain secure and be analyzed to illuminate critical insights and enable effective decision-making.
Focus on efficiency
Once agencies and their private-sector partners account for privacy requirements and security concerns, the next objective is to efficiently sort, analyze and share the data to extract meaningful value. While the amount of data available to agencies has multiplied, so too have the tools available to analyze them and provide insights. Artificial intelligence and machine learning can play important roles and, when used appropriately, save time and deliver valuable insights.
In concert with good CX practice, agencies and their industry partners should connect their analytics initiatives directly to mission outcomes to accelerate the use of data as a strategic asset throughout the federal government
Data platforms that are loosely coupled and tightly integrated can address hindrances to data sharing across agencies. This creates open architectures that allow federal employees and the public to access the information they need while putting enough controls in place to ensure the data is not misused.
Standardization is another powerful tool to accelerate data sharing. When systems are not interoperable, the transfer of information is hindered. Standardizing methodologies and technologies whenever possible can speed up and enhance these transactions. For example, DataOps, which leverages DevSecOps principles for secure data analysis, will solve inefficient data generation and processing problems by improving data quality and applying controls in an automated manner.
Federal agencies like the Census Bureau have access to massive amounts of data that will transform how the government interacts with and supports the nation’s residents. However, before agencies can capitalize on the full potential of that data, agency leaders must collaborate with one another and with their vendors to establish comprehensive strategies and procedures for effective data usage and management.
To ensure the appropriate and secure use of data, agencies must account for many competing priorities. Resolving those can be challenging, but when all parties focus on the outcome that the data serves, potential conflicts become collaborative solutions. When multiple parties work toward commonly understood goals, their unique challenges can be overcome to enhance constituent service and safety.
Evan Davis is Executive Managing Director, Business Process Services, at Maximus.
