A Project Manager's Perspective on the GAO’s Federal Data Transparency Report

September 27, 2013 Dennis D. McDonald

Introduction

This is one project manager’s commentary on the Government Accountability Office (GAO) report number GAO-13-758, Federal Data Transparency: opportunities remain to incorporate lessons learned as availability of spending data increases, dated September 2013.

The report, generated at the request of Congress, assesses whether the Federal government has learned and put into practice “lessons learned” from public financial reporting associated with 2009’s Recovery Act. The report also summarizes how the Recovery Act was implemented and how expenditure data were gathered and reported via the websites, USAspending.gov and Recovery.gov. It also describes and to some extent critiques the strategic coordination role performed by the Government Accountability and Transparency (GAT) Board with respect to standardizing how financial data and organizational identifiers are being captured, managed, and shared.

The report’s recommendations include:

For Federal financial transparency efforts to be successful there needs to be a plan developed that includes a timeline, specific standardization responsibilities, and ongoing congressional support with regular engagement of stakeholders.
Management and governance of the process needs to be beefed up. Currently the governing GAT Board is a coordinating body with strategic not implementation responsibilities. To some extent it coordinates the standardization efforts of several different agencies but it has no explicit implementation authority.
Whatever transparency standards and processes are developed, the “burden” on those outside government who may need to submit data to the government as part of it financial data transparency program needs to be taken into account.

Overall the report is clear and direct in its assessment of both accomplishments and deficiencies. On the plus side we are seeing progress towards standardization of government financial data plus useful demonstration programs. Recovery Act reporting has definitely shown what is possible in a very short time when goals are clear, governance is in place, and resources are available.

On the downside, the report points out that the current governing body — the Government Accountability and Transparency (GAT) board — manages strategy but does not manage implementation. As result the Board operates partly by “piggybacking” on various agency efforts such as those managed by Treasury. This lack of implementation authority raises governance and resource issues.

My Perspective

I am examining the GAO’s Federal Data Transparency report from the perspective of a professional project manager whose experience includes working with many different types of data in a range of public and private sector settings including manufacturers, insurance companies, public utilities, government agencies, and nonprofits.

Stakeholder Engagement

Some of what the GAO report says is needed, such as the need to engage with stakeholders, is something learned in Project Management 101: Failure to engage stakeholders is a leading cause of project failure.

An interesting part of the GAO report is a discussion of some of the creative ways one can engage with stakeholders. I call this collaborative project management. In its most basic form collaborative project management involves the use of open and transparent methods in communicating and sharing information among project staff and stakeholders.

This includes intelligent use of collaboration technologies and social media to support communication and information sharing. I believe that projects to make government more transparent, including projects that involve data, IT systems, and public access, should themselves be made more open and accessible to the public as they evolve. Using appropriate technologies to support communication and collaboration is one way to accomplish that.

Project Scope

When a project deals with data in many different formats connected with many different stakeholders, the project must clearly address both scope and standards issues.

“Scope” needs to be defined early on. You have to address the who, what, where, why, how, and how much questions when developing the project plan. You have to make decisions about “what is in ” and “what is out.” Decisions around these questions will drive the project plan, required resources, how and what is communicated, and the approach to project management. Failure to address project scope leaves you wasting time, resources, and potential political support.

What the GAO report does not describe is how the scope of the GAT Board and its activities have been defined and made operational. Answering questions like the following is important to measuring performance and success of transparency efforts:

How is “transparency” defined?
What does it mean to make spending data “transparent?”
What is “transparent” and to whom is it “transparent”?
For what purpose are the data made “transparent”?

How are Standards and Transparency Related?

Questions like the above seem to me to be fairly standard questions that a project manager, tasked with developing a new information system, will ask about the information that needs to flow through the system. For example, at times while reading this report I got the impression that the terms “transparency” and “standards” were being used almost interchangeably, as if the development and implementation of “standards” automatically leads to “transparency.”

Standardization and transparency are not the same thing. One is not always always a precondition for the other. Yes, establishing standards governing definition, terminology, data types, semantics, formats, and metadata can, when implemented and used, increase interoperability across systems and facilitate public access. The reality is that developing and implementing data standards — including data standards governing multiple financial systems — takes time, money, and planning. If you’re piggybacking a standards process on top of standardization processes already underway and managed by a confederation of institutions, you may have to compromise, starting with possible dependence on someone else’s schedule for developing and implementing standards.

Some of this is mentioned in the GAO report. Its recommendations — develop a plan, have a governance process, promote standards, involve stakeholders, implement legislation — can’t be faulted. But the report does not adequately address definitional and scoping issues. It is not clear to me if this is due to the GAO’s approach in doing this report or to the failure of the GAT Board to adequately address definition and scoping issues. As a result, the practical logistics of how to implement the recommendations aren’t at all clear to me since I’m looking at needed clarity from a project manager’s perspective.

Some implementation related issues are touched on in one of the GAO report appendices where the Department of HHS comments on the draft report. HHS gets it right and emphasizes the need to address specifics when it comes to implementation. That means, for example, agreeing on data elements and on how business processes will need to change in any move toward more standard data and processes. (See the GAO report’s Appendix III: “Comments from the Department of Health and Human Services.”)

Managing Data Projects

Changes to business processes need to be addressed as part of any implementation plan involving data and information systems. Anyone who’s ever managed a project where data have to be extracted, digitized, transformed, moved, corrected, and/or displayed and visualized appreciates the reality of working with data. This is true whether dealing with financial, structured, unstructured, graphic, textual, or all of the above data types. There are fixed and variable costs associated with each and every data element and process that needs to be touched on via manual or automated processes. These costs and processes need to be planned for, prioritized, and managed. This requires agreement on goals, processes, resources, and — above all — this requires a plan.

The need for a transparency implementation plan is addressed in the GAO report. What the report leaves out is what might happen to current transparency efforts if the recommendations are not followed. When I asked myself that question, what popped into my mind was the phrase “Doomed To Fail,” for reasons already discussed above:

No real governance or implementation plan.
Dependency on standards initiatives that may or may not be compatible with interoperability.
Failure to define goals, scope, and intermediate deliverables.
Failure to define “transparency.”

Of the above issues I think the last is the most serious. On the one hand the GAO report seems to be equating “standards” and “transparency.” On the other the focus of the GAT Board appears “limited” to financial expenditure data.

Fundamental Questions

What seems logical from a project management perspective is that, for real progress of transparency to be made, we need to agree on three important questions:

Transparency of what?
Transparency to whom?
Transparency how?

If we’re focusing primarily on financial information, then we follow a path that requires interfacing with and prioritizing a variety of departmental, national, and international financial and identifier standards. We need to understand on how these data will be standardized and how services will be developed to provide for data access, e.g., who should be responsible for data structures and data quality, who is responsible for providing access, what the role of API’s and automated tools in data access and manipulation will be, and so on.

If we’re also talking about transactional and performance data — e.g., “how many people were hired, treated, or otherwise helped as a result of the program generating these data” — we’re also raising important project scoping issues. We must involve the agencies and programs responsible for the data and the people served by these agencies and programs. Given the scope and huge variety of federal agencies and programs, resolving such complex issues that permeate all of government implies the existence of a governance body, policy, and procedures around which priorities can be established and enforced. Hopefully such issues are addressed in pending legislation such as the DATA Act.

Conclusions

There are issues that need to be addressed sooner rather than later if efforts to make government operations more open, transparent, and accountable are to succeed:

The cost-effectiveness of standards needs to be assessed. For example, are there instances where implementing standards might actually delay making data accessible to the public? How do we assess the pros and cons of such situations?
In the GAO report references are made to possibly using technology and metadata base systems to facilitate public access to non standardized data. Are there situations where it would be appropriate to consider technical approaches that bypass or delay the need for data standardization?
Making data accessible, whatever technical approach is taken, takes time and money. Given the current state of federal finances (sequestration, deficit reduction, flat or dropping agency budgets, retiring experienced work force) where will the money come from to pursue transparency?

The budget issue is the one that worries me the most. While there is no doubt in my mind that improving data transparency for both financial and non-financial data is an important and worthwhile goal, I also know based on my data management experience that any effort requiring an enterprise level shift in data formats or associated business processes can also require a substantial early peak in required resources. If in fact we need to delay implementing data transparency programs, perhaps we can use our time wisely by doing a more detailed job of research and planning.