Federal Government Leads Best Data Practices

Federal Government Leads Best Data Practices: Impacts of FEPA law especially useful to knowledge workers for next few years

Currently, all of your knowledge workers handle data, and they have all learned to handle data absent from any consistent method or guidance. In other words, data handling with most organizations is haphazard, inconsistent, and unaided. Increasing knowledge worker data literacy represents the largest untapped source of hidden productivity both inside and outside the federal government.  

A recent federal law provides an opportunity to observe the effects of increased data literacy on the entire organization. If organizations within the Federal Government increase their ability to serve the public as a result of increased data focus and capabilities, it will be time for the rest of the world to get on board with modern data practices.

Progress has been slow but steady since the passage of the 2018 Federal Data Law. Without precise numbers, it is nevertheless clear that the U.S. Federal Government is on its way to besting its private sector counterparts in data management. It will become common to look to federal data management practices for guidance, experience, and expertise instead of looking to industry for these traits. This is because it is now against the law to manage data poorly within the Federal Government. 

The Foundations for Evidence-Based Policymaking Act (FEPA) of 2018 was signed into law on January 14, 2019. The bill passed the House and the Senate unanimously.  Only one possible explanation for this kind of bi-partisanship:  FEPA must have been sold as “this is for the data folks, and don’t worry, no one will find anything in the bill to get mad about.”  In other words:  throw a bone to these folks, they all seem to want the same thing. 

One of the stated purposes of the bill was to require improvement of federal data management practices. While four years have passed, significant investments have made their way into the budgets of agencies. If these programmed funds continue to be spent wisely, most agencies will easily pass private sector practice maturity.

Key provisions of FEPA include:

1. All federal data is now open by default (except sensitive or personal data)

2. Non-political, objectively qualified CDOs with stated responsibilities are required

3. Use of open data and open models are required in policy evolution

4. Penalties for noncompliance are higher than those imposed by the Health Insurance Portability and Accountability Act of 1996 (HIPAA)

This last provision is easily the scariest as all have heard of dreaded HIPAA penalties. I will address the remaining three and point out opportunities for the private sector to learn from the government-FEPA experience. 

Federal data is now open by default

In the past, the public was required to figure out what open data was and then request it from the responsible organization. Far too often, the results were delivered via a PDF document. Now all federal organizations will benefit by making their data accessible.  While they won’t have certainty as to what data will be requested, they understand the correct defensive posture is to make their data as accessible as possible or risk spending precious resources supporting a cottage industry of hidden data factories. [Redman 2016]

Federal data practices are already on par with the private sector.  Supplementing existing requirements is the ability to quickly and efficiently deliver potentially requestable volumes of data. The only possible stance to take architecturally is to design increasing amounts of data for shared use within the organization and by external users. 

The private sector will be able to observe both the progress and success of these efforts that will take place as agencies convert increasing amounts of legacy data to largely accessible data presentation layers. The existence of the Federal CDO Council will permit evaluation of these efforts with investments guided by the new legislation.

Non-political, objectively qualified CDOs are required

“Non-political” is important in that these data leaders belong to the non-political class and should be less subject to external pressures. This relative freedom to focus on the job will be used to set standards for the rest of the community. Results here are of critical importance to the data leadership community. All studies show but few seem to comprehend that less than 20% of data-related challenges are technology based. The other 80% are people- and process related challenges requiring the use of transferable skills. The effect of trading out political abilities for domain competence will quickly become evident.

“Objectively qualified” is equally important. The CIO community has not yet settled on the right type of CIO; it has lurched from IT profession-based leadership to business-based. We still do not know the right qualities of a CIO as evidenced by their still relatively short tenure, open-third that of similarly ensconced CFOs. CDOs have also had waves of data scientists, architects, business analysts, and others, populate their ranks. No one best set of CDO requirements have yet solidified in the private sector. Objective position specifications will permit the government’s data leadership community to understand what characteristics of data leadership work. Understanding these, the community can increase the focus on the development of a body of knowledge and other required collateral.

Use of open data and open models required in policy evolution

This last provision carries the most potential impact.  It requires policy changes to follow a strict process. Consider an example:  Agency A believes that Policy #2 should supplant Policy #1. It is a violation of FEPA to change from Policy #1 to Policy #2 without specifying in advance both the model to be used to evaluate results and the open datasets that will be used. 

On the surface, the request is reasonable:  use science-based methods to influence policy evolution. Other benefits include:

• the formalization of various organizational datasets and approaches to data engineering/architecture

• The optimization of data quality engineering approaches

• increased decision-making transparency, permitting citizens to observe and replicate analyses

Challenges will inevitably arise from boundary and definitional perspectives. Indeed, one of the most asked questions of newly installed data leadership is:  what is the boundary between CDO and CIO functions? Of course, the fact that the question is asked in this manner indicates that it is being approached from an adversarial perspective. Instead, the question should be:  how do the two functions work together to best serve the organizational needs? FEPA spells out 14 specific responsibilities, and this list makes a great starting position. 

Let us watch, learn and evolve from these initial positions. Since we do not have best or even proven methods to effectively teach data handling, a lot of work still needs to be accomplished, though many agencies have stepped up to the plate. FEPA’s implementation provides an unprecedented opportunity to increase focus on data and data issues within the Federal Government. At the very least, the government should standardize data literacy education in the same way that all federal workers take annual cybersecurity training. Just this development alone will be significant. The opportunity here is that the Federal Government is going to make a major investment into data-driven government. It will take a while for useful standards to emerge, but organizations are already benefiting by increasing their data literacy.

Acknowledged Data Management authority, Peter Aiken, is an Associate Professor at Virginia Commonwealth University, President of DAMA International, and Associate Director of the MIT International Society of Chief Data Officers. As a practicing data consultant, author and researcher, he has been actively performing and studying data management for more than 35 years. His expertise has been sought by some of the world's most important organizations and his achievements have been recognized internationally. He has held leadership positions and consulted with organizations in 30 countries across numerous industries, including defense, banking, healthcare, telecommunications and manufacturing. He is a sought-after keynote speaker and author of multiple publications, including his latest Data Literacy.