Most people who work in technology are familiar with Moore’s Law- the principle formulated by Gordon Moore of Intel that the power and density of a microprocessor rate would double every 2 years.
When people mention Moore’s law, they rarely notice just how extraordinary this observation really is. Just to put it into perspective, I have heard tell that if you were to apply Moore’s law to the automotive industry, a Volkswagen would now be capable of driving at 300,000 miles per hour and cost a dime.[1]
Yet despite Moore’s law and its consequences, the government acquires and deploys complex IT systems pretty much the same as it acquires cars and office chairs. It makes no sense. Henry Ford would be pleased with our current IT acquisition system; Steve Jobs, not so much.
At a recent forum, Retired Air Force General Greg Touhill illustrated the problems of applying an antiquated acquisition system to IT acquisition and deployment. Similar to the concept of “dog years” versus “human years,”[2] “Touhill’s Theorem” postulates that if you accept that the lifespan of a computer system to be 3 years, and the average lifespan of a human to be 75 years, than 1 human year should equal 25 computer years! “Touhill’s Theorem” illustrates, in an amusing fashion, the insanity of acquiring IT in the same fashion we acquire commodities and goods in so-called “meat space.” [3]
The universe of cyberspace has fundamentally different characteristics, so it is not a leap of imagination to suggest that it should be acquired and deployed in a fundamentally different fashion.
Unfortunately, efforts to reform IT acquisition and accelerate the transition to cloud computing has a record unsullied by success. Seventeen years following the signing of the seminal Clinger Cohen legislation that laid the foundation for the federal government's acquisition and management of IT; and ten years after the E-Government Act was passed that established a Federal Chief Information Officer; program failure rates and cost overruns still plague between 72 and 80% of large government IT programs. Some have estimated the cost to the taxpayer to be as high as $20 billion wasted each year.
One commentator, in detailing this sorry state of affairs, chronicled the many attempts that have already been undertaken to address much-needed IT acquisition reform.[4] In his LLM thesis for Steve Schooner’s government contracts program at George Washington University, Micheal Garland notes that “[t]he United States government has spent over $1 trillion dollars on information technology in the last fifty years and the results have been consistently dumbfounding, often highlighted by a dazzling array of waste and duplication…..
There’s a problem here, and its hiding in plain sight: The massive fragmentation of IT acquisition, the lack of consolidation of certain processes, the internal competition between organizations, and the failure of the government to create an over-arching IT strategy are the flaws at the heart of a government-wide dysfunction that disrupts IT procurement…Until now, there has been little critical assessment of the current fractured procurement structure that hampers the government’s ability to competently spend on IT.” [5]
These issues take on particularly stark contrast when one considers the current tectonic shift to cloud computing. Cloud computing is a new name for an old concept: the delivery of computing services from a remote location, similar to the way electricity and other utilities are provided to most customers. In some ways, cloud computing is reminiscent of mainframe computing in allowing users to share the resources of a central computer, the way most users acquired computing services before the advent of the personal computer.
What is new, however, is that cloud computing is far more powerful and useful than previous generations of remote computing.
Current research shows that about half of federal agencies have adopted cloud computing in some way—21% are already moving forward with cloud adoption and 29% are in the early stages. [6]
Migration to the cloud is moving slowly for a number of reasons. The Government Accountability Office and InformationWeek Government conducted assessments in July and October 2012, respectively, of the status of Federal government cloud computing adoption. Both identified a number of roadblocks to moving services to the cloud.[7]
Recognizing the enormous potential benefits of cloud, five years ago then- Federal CIO Vivek Kundra released the Federal Cloud Computing Strategy (FCCS).[8] According to the FCCS, “[c]loud computing has the potential to play a major part in addressing ... inefficiencies and improving the delivery of government services.
The cloud computing model can significantly help agencies grappling with the need to provide highly reliable, innovative services quickly despite resource constraints.” For the federal government, “cloud computing holds tremendous potential to deliver public value by increasing operational efficiency and responding faster to constituent needs.”[9]
The FCCS noted ruefully that “the Federal government’s current Information Technology (IT) environment is characterized by low asset utilization, a fragmented demand for resources, duplicative systems, environments which are difficult to manage, and long procurement lead times.”
To help address these challenges, the FCCS sought to “[a]rticulate the benefits, considerations, and trade-offs of cloud computing; [p]rovide a decision framework and case examples to support agencies in migrating towards cloud computing; highlight cloud computing implementation resources; and [i]dentify Federal Government activities and roles and responsibilities for catalyzing cloud adoption.”[10]
The FCCS was a visionary document outlining a clear roadmap for the government’s adoption of transformative new technology. But what the FCCS failed to do was to outline a clear path forward through the current IT acquisition maze to support (rather than impede) the many benefits of cloud adoption by the government.
This is critical, because the IT/cloud ecosystem has unique attributes. It is a perishable item with a continuously accelerating life cycle. Acquisition and deployment methods spanning months virtually guarantee the eventual deployment of solutions that are obsolete- right out of the gate.
Furthermore, even assuming an expedited acquisition and deployment approach, the cumbersome and time-limited nature of the current appropriations process means that program funding will occur late in the process, in artificially defined increments and subject to strict “color of money” time constraints.
Most tellingly, agency Chief Information Officers (CIO’s) lacked the fundamental authority necessary to do their jobs. With a few, very narrow exceptions, CIO’s lacked even basic budget authority over enterprise-wide IT spending, rendering them incapable of articulating a meaningful vision for enterprise-wide IT initiatives.
Confronted with these issues, in the 113rth Congress, the Oversight and Government Reform Committee undertook a major initiative to address IT reform. The Committee developed a legislative package called the Federal IT Acquisition Reform Act (FITARA) that included the following fundamental changes to the IT acquisition process.
FITARA focused upon two key areas. First, fundamental reform was needed with the responsibility, accountability and authority of the Chief Information Officer for each of the major federal agencies. CIO’s needed to have budget authority over IT programs, because CIO’s were unable to manage their agency IT deployments because they lacked control over their agency IT budgets.
This is because since the statutory establishment of CIO positions within federal agencies in 1996, Congress and the executive branch have debated the proper extent of CIOs’ authority and jurisdiction. In private sector organizations with CIOs, this person often serves as a senior decision maker, providing leadership and direction for information resource development, procurement, and management, with a focus on improving efficiency and the quality of services delivered. But not in the federal government.
In federal agencies, the missions, responsibilities, and powers bestowed on CIOs both by law and in practice are far less clear. For example, although the CIO responsibilities delineated in 44 U.S.C. § 3506 suggest that federal CIOs are the primary officials in charge of planning and maintaining IT resources in their respective agencies, the act does not explicitly identify federal CIOs as having any budgetary control or authority over IT resources.
Second, the cumbersome, time-constrained appropriations process is fundamentally “out of synch” with the push to agile IT development practices and Dev Ops-focused IT deployments. The fiscal law constraints surrounding the process of appropriating and obligating resources to federal agencies posed a significant impediment to the timely and effective roll out of modern IT systems. Broader, more flexible budget authorities were desperately needed to accelerate cloud adoption across the federal government.
The good news is that empowerment of CIO’s across the federal government was successfully enacted in the final version of FITARA. But the second major problem- the disconnect between the appropriations process and fiscal law constraints on IT acquisition and deployment-- never made it through the Senate. Provisions in the original House-side version of FITARA included the establishment of a revolving capital fund structure to address the need for broader budget flexibilities.
Unfortunately, due to the controversial history of revolving capital funds, these provisions were not included in the final version of FITARA that was signed into law.
As a result, this same impediment continues to throttle wide-spread and timely cloud adoption across the federal government. Until this issue is squarely addressed, and the problems associated with the current appropriations/technology adoption mismatch are resolved, the benefits and promise of cloud will remain on the slow track for the federal government.
[1] Another example is an Osborne Executive portable computer, from 1982, with a Zilog Z80 4 MHz CPU, and a 2007 Apple iPhone with a 412 MHz ARM11 CPU; the Executive weighs 100 times as much, has nearly 500 times the volume, costs approximately 10 times as much (adjusted for inflation), and has about 1/100th the clock frequency of the smartphone.
[2] One “dog year” is supposedly equivalent to seven “human years”.
[3] The concepts of “meat space” and “cyberspace were first articulated in the science fiction classic “Cryptonomicon” by Neal Stephenson. The terms are self-explanatory.
[4]“Reforming IT Acquisition Reform- 50 years and Counting: Its finally time for a central strategy for government IT”. R. Michael Garland; George Washington School of Law (1/14/15), LLM Thesis (available from the author).
[5] Id. at 3.
[6] InformationWeek Government, 2013 Federal Government Cloud Computing Survey, October 5, 2012, p. 8. This survey may be accessed at http://reports.informationweek.com/abstract/104/9047/Government/research-federal-cloud-computingsurvey.html?cid=pub_analyt__iwk_20121008.
[7] U.S. Government Accountability Office, Information Technology Reform: Progress Made but Future Cloud Computing Efforts Should be Better Planned, GAO-12-756, July 2012, at http://www.gao.gov/assets/600/592249.pdf; InformationWeek Government, 2013 Federal Government Cloud Computing Survey. These include such issues as ensuring compatibility (e.g., data portability and interoperability) with legacy systems and processes; increasing the level of expertise and experience; stabilizing the standards process and obtaining additional guidance/governance to avoid cloud services “sprawl.”
[8] Office of the U.S. Chief Information Officer, U.S. Federal Cloud Computing Strategy, February 8, 2011, athttps://cio.gov/wpcontent/uploads/downloads/2012/09/Federal-Cloud-Computing-Strategy.pdf.
[9] Ibid.
[10] Ibid. In the face of these challenges, however, there are two strong drivers of cloud adoption: lowering the cost of ongoing IT operations and reducing capital investment in servers and data center equipment. It appears that in spite of existing challenges budget pressures may play a significant role in driving cloud adoption.