(See the original .pdf)

Introduction

The Commissioner of Social Security asked the panel for recommendations on the agency's plans for operation, maintenance and replacement of the many existing applications and databases supporting current business processes. Most of these systems were independently developed and they are in need of modernization.

The Future Systems Technology Panel (FSTAP) engaged key SSA officials to get a better understanding of the current systems landscape, as well as a picture of planned investments. These discussions confirmed the Commissioner's observations that the agency has a wide array of systems, developed over a long period, residing on different technical platforms, utilizing a wide range of programming languages and technologies. In many instances, these systems do not integrate well together, causing significant inefficiencies, both in business operations and in the operations and maintenance of the IT application environment.

Observations

At the heart of SSA's processing environment is the Master Data Access Method (MADAM) system. MADAM is actually a series of databases, which house the core account information for all SSA customers, and it is the data foundation that the overall systems architecture is built. MADAM was originally built in the 1980s when SSA converted its tape databases to DASD to enable online, direct access to its master files. MADAM operates in a proprietary, COBOL/ALC-based environment. As a result, the system's capabilities fall short of those of a modern relational database system. The agency is currently in the midst of a long-term strategic migration to move from MADAM to a more modern, DB-2 based environment.

An industry group called the Computer and Communications Industry Association submitted a report to the FSTAP panel suggesting that SSA pause the MADAM migration to reconsider its technical architectural and design decisions, specifically the decision to utilize DB-2 in a mainframe environment. The National Research Board raised similar concerns in a study in 2007. While the FSTAP agrees that other technical approaches might have been better choices, we do not concur with their recommendation that the agency stop its current program. The agency's design utilizes proven technology that is well accepted and heavily used in the industry. Given the substantial investment and clear progress SSA has achieved in 2

the current program, the FSTAP does not recommend stopping to consider other technical alternatives at this time.

However, the FSTAP has concerns about the long-term ability of the new architecture to support modern, web-based technologies in the future. We recommend that SSA consider bringing in outside expertise to conduct a review of the data model and schemas, as well as the surrounding applications, to achieve the maximum degree of openness and flexibility to allow the new data architecture to support future service delivery requirements. In addition, the agency should develop a plan to evolve access to those datasets necessary to meet the changing needs of the agency and its customers.

In addition to the MADAM conversion, SSA has a number of other strategic application development projects underway, including a redesign of the Disability Processing system, the development of web-based applications to interact with the public, and the agency's vision for a "seamless" processing system.

Based on information SSA officials shared with the FSTAP about the agency's strategic objectives and plans, we believe that the projects currently underway seem to align with the agency's current strategic business priorities and plans.

Recommendations

The panel recommends that SSA consider developing a comprehensive agency-wide strategic systems development roadmap. The focus of the roadmap should be on critical strategic systems and applications that will have significant impact on SSA customers, the agency's business costs/results and/or impact on IT costs for operations and maintenance. Give priority consideration to those systems, which provide foundations for further development, such as the MADAM migration.

The strategy should differentiate those systems that clearly must be updated, with those that should clearly remain in their current state, and a process for evaluating and making decisions regarding those in the middle. In addition, the strategy should include fundamental architectural decisions regarding processing platforms, storage technologies, programming languages, etc. Once decisions are made, processes should be implemented to ensure that all new development activities comply with architectural standards and directions.

Resource requirements for both new development and ongoing operations and maintenance should be factored into the planning of the roadmap. The strategy should include a high-level view of a realistic future state, along with a plan that outlines the major steps the agency will take to achieve its vision.

The panel believes it is critical that appropriate stakeholders are fully engaged in formulating the strategic systems development roadmap. Stakeholders include the 3

IT community and key business leaders throughout the agency. Engaging business stakeholders is essential to ensure that the systems migration strategy is aligned to business priorities and plans. Business buy-in is also critical to maintaining the discipline to ensure that investment decisions are driven by strategic priorities over both the short and long term. Both IT and business stakeholders should be brought together in a comprehensive governance process to guarantee progress toward the agency's strategic systems development vision.

Managing the demand for IT services is essential to the long-term success of the strategy. SSA should have a process to ensure that its limited resources are invested to address the highest priorities and requirements to meet the mission of the agency. Business stakeholders must understand and support the overall strategy and agree to focus their service and development requests on those activities that will advance the organization toward realization of its strategic vision. Processes should be implemented that will fully engage stakeholders to ensure that strategic priorities are used to drive IT investment decisions.

Final Note

At the November FSTAP meeting, the CIO shared some of the initiatives that he has undertaken since arriving at SSA. The FSTAP is pleased to hear that work is well underway to implement a number of our recommendations. In addition, the Panel has volunteered to assist the CIO develop and implement its strategic systems development vision and roadmap.