Case Study of Public Safety Networks pt. 2

28, 2013

If we know that legacy systems generally reduce end user satisfaction, cost extra support dollars, and limit our organization’s ability to increase scale/scope, then why do they persist?  Why do legacy systems persist in state-level agencies? Why do they persist in law enforcement organizations? Why, if a recent UNISYS/MeriTalk survey is fully representative of the current situation, do federal agencies spend almost half their IT budgets (to the tune of more than $35B annually) on legacy systems?[1]

Factors triggering legacy effects

There is no generally agreed upon definition for when a particular IT device, application program, system or overall IT architecture crosses into the “legacy” category.  However, like some other phenomena, “you know it when you see it.”  Legacy systems are typically classified as such based upon their effects on organizations.  The IT “legacy effect” is a condition when necessary upgrades and maintenance are cost prohibitive (time and resources) due to the constraining influence of the installed base of existing technology and operational practices.

Legacy systems often result from well-intentioned efforts to increase IT architecture customizability and stability. Customizability here means that systems are designed and developed to meet specific needs of local users.  Customizability in general is a good thing, but maximizing system capabilities for local needs sometimes sub-optimizes more global objectives. For example, if a local law enforcement agency builds a stovepipe database of crime hotspots then the local needs of officers might be met, but data sharing with other agencies will be limited and application upgrades and ongoing system support might be cost-prohibitive. Since state and federal grants often target new development (vs ongoing maintenance) the incentive to build a new one-off solution is great. However, the result can be a complex patchwork of information systems that not only fails to meet the changing information requirements of law enforcement, but also is more prone to failure and security breach (fragility of the “Frankenstructure”).

Stability here means the IT architecture’s ability to absorb new requirements (scale or scope) without adjustment and maintain reliability over time.  Stability is also characterized by words like “robustness” and “resilience.” Stability also reflects a situation where IT protocols, practices, and policies are enforced across an organization in order to limit the volume and variety of technology elements that comprise its IT architecture. While customizability would tend to increase IT architecture complexity, stability would tend to reduce IT architecture complexity.  However, the result can be IT architecture rigidity that also fails (for different reasons) to meet the changing requirements of public safety professionals.

Overcoming legacy effects

So how does Nlets successfully overcome the “legacy effect” for its member states? Organizations focused on overcoming legacy effects need to address two sets of requirements - bootstrapping and adaptiveness[2]Bootstrapping reflects the challenge Nlets faces in getting end users to use next generation technology for the first time.  User participation, training, authoring white papers to aid in the business case justification, and distributing federally-sourced grant funds are all part of helping member states to implement (stand up) modernized systems capabilities. Adaptiveness reflects the challenge Nlets faces in deploying systems that allow for cost-effective, schedule-efficient modernization when new functionality is required in an existing system. Providing expertise in on going data and network security, public sector XML standards leadership, and working with “leader” or early adopter states as proof of concept pilots are all part of helping member states to implement upgrades and revisions to existing systems so the “legacy effect” does not create IT fragility or rigidity over time.                     


Modernizing legacy systems is hard work, but worth the time and attention. Legacy systems persist for “good” reasons, so thinking up front about the bootstrapping and adaptiveness requirements will help overcome the fragility and rigidity outcomes that older, existing systems can generate. At the heart of the matter are well-intentioned IT objectives.  Focused attention on balancing the tension between customization and stability needs has been a successful approach for Nlets, and can be used by other organizations striving to keep their IT architectures relatively modern…unless of course you like spending half your IT budget on your “Frankenstructure.” 

About the blog series:

This entry is part of a series of entries on a National Science Foundation funded research project examining public safety networks in the United States.  Nlets was one of several case studies and we focused attention on it as a result of a tip from a Curt Wood of Massachusetts. Additional entries in the series will cover a variety of topics including – dealing with assessing the success of networks, legacy systems modernization, the advantages of being a network organization, challenges Nlets has faced, and some challenges Nlets faces going forward.

About the authors:

Martin’s research examines the formation, operation, and evolution of IT-enabled public safety networks.  His expertise involves designing IT (from applications to architectures) that supports public safety inter-agency data sharing and collaboration. He is currently an Assistant Professor at Northeastern University in Boston, MA.

Dax is broadly interested in the management and use of technology in organizations, particularly in the public sector. More specifically his research focuses on how governance influences technology and technology influences governance as they interact over time. He is currently an Assistant Professor at California State University Channel Islands in Camarillo, CA.


[2] Source: Hanseth, O. and K. Lyytinen (2010). "Design theory for dynamic complexity in information infrastructures: the case of building internet." Journal of Information Technology 25(1)