John Thomas Flynn's California Network Nightmare

In the battle for the RJ11 jacks of California state workers John Thomas Flynn never had a chance.

In the battlefor the RJ-11 jacks of California state workers, John Thomas Flynn never had a chance.

Flynn, the state's new chief information officer, declared earlier this year that the California state government would officially exit the business of owning and operating networks for state agencies. Instead, it would seek a commercial partner to solve "the business problems of state agencies" and provide for the "rapid deployment of new technologies to meet the state's ever-increasing telecommunications demands."

On the books, the decision to privatize the state's $300 million telecom works is a no-brainer. California agencies are running on no fewer than 17 separate networks, which siphon off valuable traffic from Calnet, the state-owned and operated voice network. Established a decade ago as a way to get telecom prices down, Calnet is now sitting on a mountain of debt, estimates for which range from $20 million to $30 million.

"We've had continued financial losses associated with Calnet," said Lee Kerscher, the state's deputy director for networks and telecommunications. "Suffice it to say, the legislature when establishing Calnet expected to achieve significant savings, and those savings have not occurred."

To add insult to injury, the state is not satisfied with the service it does get from Pacific Bell, to which it pays a hefty $87 million annually in assorted service fees. "We're the largest customer of Pacific Bell, and we're not getting the kind of service we deserve," Flynn said.

To untangle the situation, Flynn says he will make the market compete for his business. "The public sector is moving in general toward consolidation and outsourcing," he said. "I hope the private sector will now step up to the plate and come forward with solutions that take an `all-or-nothing' approach. Let the private sector put together consortiums and offer true public/private partnerships and long-term service agreements."

Flynn's plan calls for the state to take a number of steps to stop the hemorrhaging. To reverse Calnet's voice traffic drain, the state has announced that Calnet will be made mandatory for all state agency voice traffic. To begin consolidating data traffic, California has awarded a contract for long-distance data transport services and will compete a master service agreement for frame-relay services and network management. State agencies have been asked to move all their traffic onto these networks once their current contracts expire.

"This will lead ultimately toward privatization of both voice and data services," Kerscher said. "The reasoning behind this is that the state is not able to respond as quickly to changes as private industry does. There's difficulty in establishing new infrastructures."

The long-term strategy for state telecommunications services will move California from a model of state-owned and operated redundant networks to a common, integrated, state-managed but privately owned and operated network. The state is aiming toward a gradual migration to improved data, voice and video services, including the introduction of high-speed Asynchronous Transfer Mode and Synchronous Optical Networks.

What Went Wrong

Calnet was positioned as an optional telecom service provider to state government institutions, including community colleges and universities. At the time, the state wanted to be able to leverage its buying power to offer services and prices that agencies could not find on the open market.

But the downside was huge. "The problem is principally around the Calnet voice environment," Kerscher said. "The state expected to get into the network business to obtain substantial savings. Several years later, we haven't saved a lot of money, and we have an infrastructure that's not suitable to the technical demands being placed on it."

According to an analysis by the state's Department of General Services, the strategy was all but doomed. Calnet was set up to produce price and service advantages, but there was no clear mandate for agencies to use the network. In fact, the decision to jilt Calnet often had little to do with price: "Whenever a choice is offered, agencies may decline services for reasons that have little to do with the prices and quality of service offered by the state provider," the analysis said. "Chief among these reason is distrust of an oversight agency."

Other flaws in the original plan included a lack of flexibility to alter pricing to fend off competitors; a politically weak policy and enforcement apparatus; marketing and sales that "cannot match the persistence of private-sector efforts"; and a human resources system that "limits the state's ability to readily recruit and retain highly skilled, experienced and well-paid telecom managers from the private sector," according to the report.

"We would go a step further and state that the concept of having a state-owned network competing for state agencies' business constitutes a prima facie thwarting of the primary advantages of establishing a state-owned network in the first place," the report continued. "These advantages are leveraging the state's aggregate buying power and eliminating redundant agency networks."

Nationwide

California is one of several states looking at blanket network privatization. North Carolina, Texas and Oregon have already chosen to divest themselves of their current network infrastructures or to "shrink down" to contractor-provisioned networks while maintaining a small, highly trained telecom planning staff. Indeed, a 1994 survey of the National Association of State Telecommunications Officers found a majority of states planning to lease, rather than own, their voice and data networks in the next five years.

"As quickly as the industry is evolving, for a state to make investments in infrastructure really could be devastating," said Carolyn Purcell, head of the Texas Department of Information Resources, which manages some rural switching centers but contracts the lion's share of its transmission needs to local telephone carriers. "Basically, privatization is the way to take advantage of the newer technology," Purcell said.

In Massachusetts, where Flynn served as CIO before heading to California, the state was also moving toward data center and service consolidation. "In Massachusetts, we were dealing with seven long-distance companies, but by consolidating to a prime vendor, we were able to save about 40 percent, representing millions of dollars, and we're doing the same in California," Flynn said.

While the pitfalls of state ownership of telecom facilities might sound self-evident, there are advantages, including management control of the facilities. Even so, analysts say the risks are dangerous. "Some states that have established a telecom network find that maybe they don't get what they expected in terms of efficiency and improvements," said Milford Sprecher, program director with IDC Government, a Falls Church, Va., government market research firm. "In California, they're not consolidated, and it turns out to be pretty costly.

"Unless you have enough traffic to use the network all the time and use all the capacity, it can be losing money," Sprecher continued. "Besides, can a state market well internally? Will other state agencies go somewhere else? Although they have a number of things they bring to the table, in terms of managing their network efficiently, it's not their business."

Hold Onto the Baby

Whatever a state's or city's decision on whether to keep title to its telecom infrastructure, public-sector executives caution that governments should take care not to cede total network management and oversight to outsiders, no matter how seductive the offer.

"When you're aligning networking technology to your business needs, strategies and missions, that has to be done by your own folks," said Otto Doll, chief information officer for the state of South Dakota. "Some people try to get an outside vendor to do it from A to Z. But from a strategy standpoint, planning should be done in-house. While a tactical plan could be carried out by a winning vendor, they will still need to be driven by the state," he said.

Doll, whose entire state telecom budget is $16.5 million, runs a principally in-house state voice and data operation. "We're considering our options about outsourcing, but the bottom line is that the cost dynamics are different for South Dakota and California," he said. "California is in a very lucrative marketplace for private telecom, and although we do have US West here, the majority of the state is serviced by 60 [local exchange carriers] and US West."

On the other hand, "it's very cost-effective to staff in South Dakota," he said. "My general philosophy is that things that deal with operations are good candidates for outsourcing because you burn a lot of time managing and maintaining networks. On the other hand, in South Dakota we spend more money on [local-area networks] than [wide-area networks]. Our people-dollars amount to only half a million, so we have to grow a lot of our own expertise. But that's not to say that we can't run a really good shop from the inside."

This month, California CIO Flynn will get a chance to test the mettle of outside vendors and consultants, who will tackle a much more complex telecom infrastructure. "It's like selling a house with a mortgage," Flynn said. "We assume we'll come out ahead. We've got a lot of companies who want to get into local phone business but are having a hard time working out agreements with local service telecos. Here's an opportunity for long-distance companies to get into local services. And it's an opportunity for PacBell to prevent them from doing it. We think we should have a salable commodity here." The job is expected to be awarded by early 1998.

Arielle Emmett is a Wallingford, Pa.-based free-lance writer who specializes in computer and network technology. She can be reached via the Internet at arielleem@aol.com.

NEXT STORY: Fedwire