Debt consolidation is stuff of late-night advertising -- offering hope to those with upside down finances. The danger is consolidation often treats symptoms of bad habits without making changes to avoid getting back to the same mess, or worse.

Consolidation in public-sector IT offers the same promise and pitfalls, and is often pursued in the name of cost savings. As a practical matter, it has at least proved useful in getting one's arms around the annual IT spend.

In California, annual technology investments could be anywhere between $2 billion and $4.5 billion, depending on the estimate. In August, Gov. Arnold Schwarzenegger issued an executive order for a comprehensive data center consolidation to help narrow that range.

Delaware, Florida, New York, North Carolina and Oregon also embarked on consolidation efforts, and Michigan reports being on track to realize $50 million to $70 million annually through consolidation. Virginia hopes to save $100 million through massive IT restructuring.

Like spring-cleaning, there is something iterative and cyclical about consolidation, talk of which scares people because it changes who controls how much -- classic bureaucratic indulgences over turf and control the public treasury can no longer afford. In September, Connecticut Gov. Jodi Rell suspended state IT consolidation less than a month before the change. Connecticut also reversed its late-1990s outsourced consolidation plan in the 11th hour. Rell says consolidation can wait until Rock Regan's successor is found -- perhaps the final indignity in the most painful CIO departure in recent memory.

Consolidation often is used as a synonym for centralization. It is not. Nor is it a new idea. Many in public-sector IT have scars to show from the last time they took up such efforts, having witnessed the pendulum swings of IT organization. At one extreme, centralization gains traction as a reaction to profligate spending and unnecessary duplication among independent agencies. At the other extreme, decentralization is often a reaction to poor service, inadequate capacity and lack of trust in a central provider.

While all centralization requires consolidation, not all consolidation requires centralization. Shared services organizations provide a middle ground for redeeming consolidation by blending the best attributes of the poles while mitigating their downsides.

In a shared services environment, governments retain internal control of core functions while maximizing cost efficiencies, bringing together frequently duplicated functions from various organizations and offering those services more efficiently at lower cost. Despite their commonsense appeal, shared services are hard to execute well because they embrace contradictions:

•balancing needs of the larger enterprise with those of agencies to act autonomously;
•competing to be the cost-recoverable provider of choice with a customer base characterized by mandates and reliance on appropriations;
•reconciling the political expectations of public officials with service-level expectations of agencies; and
•supporting enabling technologies while embracing disruptive technologies.

Among disruptive technologies, the enterprise service bus is a natural extension of consolidation with its emphasis on maintaining local control and autonomy in applications, while connecting to a larger integration fabric of shared services.

That said, disrupting old, tired processes and old-line organizations through IT consolidation and shared services is only prologue. The important story comes next -- public institutions that are structurally broke and operationally broken. In The Price of Government, David Osborne and Peter Hutchinson make a compelling argument for the need to consolidate government functions around a few key priorities. Public entities must resist the temptation to make across-the-board cuts and focus investments on a prioritized list of outcomes citizens expect.

Delivering against those consolidated outcomes is the sweet spot in public-sector IT during the opening decades of this century. To that end, let's disavow the language of minimal expectations and failure. We will not get there by "doing more with less," but only by doing things differently. KW