Analyst Recap: SLED has legacy, too!
Published: May 26, 2016
A fresh report from GAO highlights the Fed's multi-billion dollar issues with legacy systems. However, a recent panel at NASCIO's mid-year conference put a spotlight on the issue from a SLED perspective.
The GAO report, "Federal Agencies Need to Address Aging Legacy Systems" (May 2016), is premised upon some interesting assumptions:
The federal government spent more than 75 percent of the total amount budgeted for information technology (IT) for fiscal year 2015 on operations and maintenance (O&M) investments...Such spending has increased over the past 7 fiscal years, which has resulted in a $7.3 billion decline from fiscal years 2010 to 2017 in development, modernization, and enhancement activities.
Now, it is to be expected that any government IT shop will be spending 65% to 85% of its budget on O&M. So, the fed is not out of line here. The crux of GAO's concern seems to be that this legacy based O&M spending is crowding out development, modernization, and enhancement (DME) activities. This is probably to be expected when top-level decision makers put downward pressure on spending without providing the additional resources for transitioning off of legacy systems.
A May 5th session at NASCIO's mid-year conference, "The Evolution of IT Systems: Legacy Modernization," addressed the same concerns from the perspectives of the states of Washington, California, and Michigan.
- Mike Cockrill, CIO, State of Washington, reported that his state spends roughly $1 billion per year on IT and that it would cost up to $2.8 billion to replace all of their legacy systems. The state has issued a comprehensive plan for legacy systems.
- James McFarlane, Director, DTMB Agency Services, State of Michigan, indicated that legacy systems are a $1 billion problem for his state. The state has a quarterly portfolio report that tracks progress on legacy migration. The legislature has appropriated $47 million to $65 million per year to support the effort.
- Christopher Cruz, Chief Deputy Director, Operations, State of California, said that his state is just beginning the process of inventorying legacy systems and developing a migration plan.
Cockrill said that Washington uses a rigorous system to classify a system as "legacy" only when it meets at least one of the following criteria:
- The system is not easily updateable due to complicated or indecipherable code, fragile interfaces or lack of useful documentation.
- Maintenance or modification of the system depends on expertise that is hard to find or prohibitively expensive.
- The system depends on software no longer supported by the vendor.
- Other risks identified by agencies, such as vendor instability, lack of alignment with enterprise architecture or lack of “bench depth.” (See PDF p. 5 at the report link above.)
In the end, the three agreed on two fundamental assumptions that must be made for a successful legacy migration policy. First, the transition period will result in increased spending as operations are migrated from the leacy to a modern system over several years. Second, the governance process for the migration must be overseen by business leaders, not "IT people," so that risks are understood and the prioritizations made by the business owners throughout the process.
In fact, one of the speakers joked that the idea that legacy migration will save money in the short term is a classic example of "execu-think" that must be corrected as early in the process as possible. Two systems will have to be run concurrently during the transition period--especially for those serving mission-critical operations.
The picture painted by the panelists leads this analyst to believe that even an aggressive effort to migrate off of all the existing legacy systems in a state would be a decade-long (or longer) process requiring significant investments. And, who knows how many currently modern systems would graduate into legacy status during this time?
As with most things in IT, legacy modernization is a journey rather than a destination. You can be sure that any modernization effort that has not secured a major boost in funding (maybe 20% or 30% above new system costs) to cover migration is either A) not a serious effort or B) doomed to failure.