Government needs to get serious about IT modernization
- By Michael Garland, Gaurav "GP" Pal
- Jun 18, 2019
A review of agency IT budgets over the last 25 years shows the U.S. government has spent over $1.5 trillion on IT since 1994. In recent fiscal years, the federal IT budget has been over $90 billion annually. Only six of the 24 CFO Act agencies have total budgets larger than the aggregated IT spend of the government. To put this in perspective, one year's IT budget is large enough to fund three State Departments.
Yet while the federal IT budget is huge, it is far from monolithic. Spending is fractured and distributed like large crumbs across all agencies. And therein lies the problem. With all the money flowing to agencies in comparatively small increments, American taxpayers, Congress and even the White House lose visibility of the spending in a way that's not commensurate with its importance. As a thought experiment, imagine if the State Department was three times its current size and had no cabinet level person to lead it, little oversight and limited individual accountability.
Meanwhile, the government is hemorrhaging good money, trying to hold together antiquated IT systems. According to GAO and OMB's own 2019 budget document, it's a known fact that over 80 percent of all IT spending is on the Operations and Maintenance of "aging legacy systems, which pose efficiency, cybersecurity, and mission risk issues, such as ever-rising costs to maintain them and an inability to meet current or expected mission requirements." O&M is over $64 billion a year -- the equivalent of more than three NASAs and more than seven Commerce Department, annually.
Given this dismal reality, the government needs to get serious about technology modernization. Modernization ends the wasted money spent on O&M and has the potential to deliver far more efficacy to American taxpayers at a much lower cost. One only needs to remember the inexorability of Moore's law, which states that computing capacity doubles about every two years. This accounts for why an iPhone is more powerful than the computers that put a man on the moon, yet even teenagers in developing countries can now afford to own them. When money is spent on operating old technology, the government does not get the benefit of Moore's law.
Unfortunately, despite bountiful evidence of wasteful IT spending, Congress has yet to get serious about retiring the nation's antiquated IT. It its most recent attempt, Congress appropriated only $250 million for technology modernization to be spent across 2018 and 2019. Even worse, it appears only $25 million actually made it into the 2019 budget. In point of fact, $25 million for modernization is microscopic when compared to the $90 billion IT budget. It's less than three hundredths of a percent (.000277). With a modernization investment like this, can the government be serious about this problem?
It's possible we've been looking at this the wrong way. All legislative efforts at reform have been structured like the IT budget, easily fractured and lost like crumbs, as the focus of IT policy has been around forcing compliance at the component level, rather than putting an emphasis on the entire government enterprise-wide. What's needed is the elevation of IT. In particular, there's an immediate need to launch the equivalent of a Marshall Plan to retire old IT debt and to replace as much as possible as quickly as possible with modern commercial solutions.
To get started, it would be prudent to appoint an IT Modernization Czar empowered to advocate for everything needed at a government-wide level and to answer to Congress. We need to task the Modernization Czar for targets and goals and to provide accountability toward progress. We need to use return on investment strategies to size, allocate and prioritize modernization budgets in light of evolving agency missions.
By analogy, too much of the government's current IT is similar to a gas-guzzling car of the 1970s, impervious to the improvements provided by innovation, missing embedded safety features, and way too expensive to keep on the road. We can keep patching it up – or we can bite the bullet, make new investments, and begin to gain the benefits and economies of modern computing. Over time, modernization will save money and significantly improve mission performance.
Upgrading and retiring expensive legacy systems can be done. Given the hard work of many dedicated and passionate individuals, we already have some of the key components needed to drive a Marshall plan.
1. Funding for Modernization. The Technology Modernization Fund (TMF) should provide seed capital to jumpstart new and innovative modernization programs. Just as private sector venture capital is a key ingredient to help new and innovative solutions get off the ground, funding for TMF should be radically increased to ensure that major modernization efforts have appropriate resources. TMF should use transparency as a means to drive accountability. A well-designed TMF program and website with detailed information on spending, utilization and successes will ensure that projects can be monitored and that progress is reported. Most importantly, Congress needs to provide serious money and serious accountability. Modernization needs to be an important high-profile priority, backed by significant investments and unrelenting accountability.
2. Special Forces for Modernization. The White House's Office of American Innovation established the Centers of Excellence staffed with industry and government experts in modernization. Through an interagency agreement process, agencies now have access to the experienced and qualified COE experts who cover a broad spectrum of modernization topics and areas.
These are fast and nimble teams that deliver results in short order. They are not incented to create long-running engagements but are focused on quick wins. This program needs to be accelerated and expanded. The discovery of opportunities for shared services should be rewarded. If one agency is building a modern forms digitization service, can it be reused and shared across government?
The COE should be directed to focus on two things – quick targets for modernization, and the potential for reuse and sharing across the government. A fortune could be saved when common modern back office solutions could be shared between agencies.
3. Smart Policy Guidance. In most cases, the task of modernizing IT involves rebuilding or moving compute workloads to modern cloud environments. OMB's recent CloudSmart policy is a step in the right direction. But the policy must be embraced, and the entire government needs to pick up the pace.
There is nothing wrong with current policy, it just doesn't go far enough. What's lacking is a willingness to significantly increase the urgency and to provide congressional commitment to make modernization happen rapidly across the entire government. With the proper focus, it should be possible for every agency to list their top 10 worst O&M IT problems, with a defined deadline and budget to get them replaced and retired. In this way, we put a stake in the ground and incrementally bring the entire government up to speed by published milestones. Ongoing security concerns should also be tackled as part of the modernization process, with an eye toward reducing security inhibitors and barriers that slow progress.
The key ingredient to make this modernization Marshall Plan a reality is to appoint a czar. A properly empowered leader could deploy the tools already available, apply some elbow grease, and scale them up at a measurable and definable pace. Think of the modernization effort as similar to the Data Center Optimization program but on steroids with an empowered task master.
The government has both the knowledge and the people to do this. There are significant thought leaders who have driven commendable modernization. Unfortunately, it has never been done at scale, with the resources required, and with empowered leadership to move the entire government enterprise forward. There is really no reason for Americans to continue to pay for antiquated and expensive IT, when less expensive modern solutions will save money, significantly improve mission performance, and create a more digital citizen-centric experience.
The time is now for Congress to make IT modernization a reality.