The Region

Interview with Carl E. Powell

Published December 1, 1991  | December 1991 issue

Carl Powell, director of automation resources for the Federal Reserve System, does not mince words in the following interview when he describes the upcoming consolidation of some Fed services, calling it "one of the most dramatic changes that the System has ever attempted to make."

In 1992, the Fed will start to consolidate some of its data processing operations that are now performed by each of the 12 district banks. According to Powell, this move will not only save money for the System but will also streamline operations and provide better services for financial institutions.

Powell assumed his new position earlier this year after serving about five years as first vice president of the San Francisco Fed, and he says he is excited to play an important role in the Fed's consolidation efforts. "The enormity of what we're trying to do with automation consolidation, and what I perceive to be the tremendous benefits for the System if it's accomplished well, was more of a challenge than I could walk away from."

Following his studies at the Stonier Graduate School of Banking at Rutgers University, Powell began his career with the System as a budget analyst at the Kansas City Fed in 1970. He was promoted to operations officer in 1976 and assistant vice president in 1978, moved to the San Francisco Fed in 1979 and became first vice president in 1986.

Region: To begin, please explain the genesis of automation consolidation for the Federal Reserve System. When and why did the idea develop?

Powell: Actually, it started through the efforts of a lot of the individual Reserve banks. Over the years, Reserve banks have looked for ways that they can improve their own efficiencies, lower costs and improve services. The Minneapolis Fed has done a lot of that over the years. You [the Minneapolis Fed] had some early discussions a few years ago with the Chicago Fed about the possibility of doing some joint processing, for example. A group was formed in about 1987, where representatives from the Reserve banks in Minneapolis, St. Louis, Kansas City and Chicago began talking about some automation consolidation for those four districts.

That, I believe, was the first formal action toward a multidistrict effort. From there it really spread—if there are some possibilities to do it for four, why not do it for 12? Conceptually, it makes a lot of sense. We operate the same basic businesses, with a few differences; but we are redundant in a lot of ways and need to do whatever we can to improve our services and our reliability.

It's a concept that we've seen a lot of other companies do and, as I said, we've done it our own way, over time, in lesser steps. It seemed like the time had come to look at the idea again and see whether there was a way for us to do it on a System level. The studies showed that it was feasible and that there could be some substantial savings.

Region: How much money does the System hope to save by consolidating services? Regardless of the amount, where will savings come from?

Powell: We don't have a firm figure at this time, [interview was conducted in mid-September 1991] because we have not completed all the planning that will tell us exactly how we will consolidate, what the transition plan will look like, what the timetable will be, and what some of the cost components will be. It's going to be in the tens of millions, maybe hundreds of millions of dollars over a period of time, but I don't have an exact figure.

The unfortunate part is that a lot of the cost savings will come in the form of people savings, which has the potential to be a very disruptive and painful process, I'm afraid. It always is when you have to downsize staff. We will save money in hardware and some of the peripheral categories associated with the data centers themselves, but the real bulk of the savings will come in the people side. That's unfortunate because it will mean that each district will be downsizing its staff levels.

Region: Have plans been made to ease the pain associated with that down-sizing, either Systemwide or by Reserve banks?

Powell: The System has constructed a fairly comprehensive incentive program that we will use to retain employees through the transition period. And each district is also constructing its own separation package plan. So, for those employees that do need to be let go, there will be some incentives to keep them through the transition as it relates to their specific skills; and, once they are displaced, there will be a separation program provided. That program will vary from district to district and will depend on the normal practices of each district, but the concept has been approved through all levels within the System.

And, obviously, as we have the capability, each district will try to absorb as many of those employees into other positions—as much as possible. Hopefully, that will take care of the bulk of the employees, so the real outplacement impact on the employee population will be as low as possible.

Region: Dallas, Richmond and New York's operations site in East Rutherford, NJ, were chosen as the three main consolidation sites. Why were three banks chosen—as opposed to one, two or four, for example—and why those sites?

Powell: The reason we chose three is based on our consultant's recommendations. They constructed a model based on input concerning our capacity needs, our current requirements and also information concerning our mature business environment five years from now. Based on all of that information, their model came up with a three-location decision; and, quite honestly, the reliability factor was a big reason for the consultant's ultimate decision. The plan is that all three sites will be redundant backups for each other, so that there will be a two-tiered backup system for each location. Today, as you know, we have Culpeper that backs up 10 facilities, and only currently has the capacity to back up one at a time. We didn't want to get into a situation in consolidation where we had a one-on-one relationship, we wanted to have at least two paths of contingency reliability for each center.

We went through a process using an external consulting company to look at site selection. We engaged IBM to assist us in the overall study. They, in turn, subcontracted with an organization called PHH Fantus in Chicago—they're a relocation company that does these types of studies.

In discussions with them, we decided we would take a three-tiered screening approach to selecting the sites. We began the process with 39 potential sites, which included all 12 main offices, all branches, the Culpeper contingency site and the operations site in East Rutherford. The first-tier screening consisted of three components: (1) susceptibility to natural disaster, (2) cost and stability of the power supply and (3) availability of labor. For each of the three components we compared each location relative to each other, that is, we didn't set any standard by which to measure them, we just rated them and ranked them based on relative merits.

Each of the components was assigned a rating, from "most preferable," to "preferable," acceptable" and "not acceptable." During this first screening any location that received a "not acceptable" rating was eliminated from further consideration. For the first component, susceptibility to natural disaster, for example, we looked at earthquakes, hurricanes and tornadoes. The analysis was based on historical data and weather maps from the National Weather Bureau. Through this process, some places—like San Francisco, which would obviously get a "not acceptable" due to earthquakes—were quickly eliminated.

The second component was the cost and stability of the power supply. Obviously, the availability and reliability of power is a very critical component of running a data center. The consultant looked at historical information in terms of outages, and also looked forward by analyzing the plans of all the utility companies for expansion for the next 10 years.

The third component in the first-tier rating was the availability of labor. They looked at data processing labor pools and general economic labor availability in all 39 locations.

So, based on the three components of the first-tier screening, we went from 39 locations to 12. We then did a very quick and preliminary review of space availability and immediately eliminated five more that we just knew would not have space to accommodate our needs.

Of the remaining seven, I then did a more detailed analysis of space availability, and asked the consultant to do a tier-three screening which included such components as cost of living, availability of housing, educational systems—the more "quality of life" components. During my further analysis of the space availability three more were eliminated. So we really ended up with four viable candidates, obviously three of them were Richmond, East Rutherford and Dallas. The fourth was Atlanta.

Dallas and East Rutherford have a natural advantage as far as facilities, because they're both in the process of building new facilities. And looking at the facilities of both Richmond and Atlanta, and looking at the time-frame in which we needed facilities, Richmond came out better. Although Atlanta was very, very strong in a lot of the other components of all three tiers of analysis, the space limitations there, and the timeframe in which we needed space, made Richmond a better choice.

Region: In what manner will consolidation occur, that is, will each consolidation site absorb all the operations of a certain number of district banks, or will each site assume only specific operations? For example, will Dallas have all securities, Richmond all ACH, etc.?

Powell: We don't know, at this point, exactly how that will occur. We're in the process of working with IBM on a transitional approach, and they're going to recommend an approach to us, at which point we will do some analysis of our own to determine the best plan.

My guess is that there will probably be a mixture of both of those scenarios. There will likely be some business or applications that may be moved a district at a time; there may be other cases that we move by business or application.

Region: What will happen to existing mainframe computers at the district banks? Will the banks even need mainframes anymore?

Powell: If we do our job correctly they won't. The whole concept of automation consolidation is to eliminate the need for mainframe processing at all but the consolidated sites. The plan is that any application that currently runs on a mainframe at any district will be migrated to one of three consolidated sites. Eventually, it will include the operations that are run at the Board of Governors. So, once we're finished with consolidation itself, there should be no need for mainframes to remain in any of the 12 districts or Culpeper, and they will be eliminated entirely.

Region: Were there parallels for consolidation in private industry that served as models for the Fed plan? Are those models successful?

Powell: I don't know that we've necessarily looked at any particular model, but certainly automation consolidation has been accomplished by a lot of different organizations in the private sector. Many banks, for instance, have been going through this process for quite some time and will continue to go through this process. A lot of it has been spurred by the mergers and acquisitions that have taken place over the last few years. As you get disparate organizations coming together under one umbrella, they've used consolidation to migrate those businesses into a single organization.

Region: Are there parallels in federal government?

Powell: There are many centralized processing systems in government. I'm not familiar with all of them, but I know that the Internal Revenue Service, for example, has some centralized, consolidated data operations systems.

I think a good parallel that everyone in the Federal Reserve is familiar with is the Treasury Direct system, which Treasury implemented a few years ago through the Federal Reserve Bank in Philadelphia. That's a system that's run out of a central location, but it services all 12 district locations that are involved in selling Treasury securities. In some ways, the Treasury Direct System could be seen as a model for what we're trying to accomplish—it works extremely well and Treasury is very proud of it, as they should be. It's a good model for us to follow.

Region: Ten years from now, how do you envision the status of consolidation in the Fed System? Will new operations be consolidated by then?

Powell: I can't really predict what will happen 10 years from now; I'm not quite sure what will happen tomorrow sometimes. I guess I would have to assume that if this effort is successful—and I'm certainly going to do everything I can do to make it successful—it will be judged on its merits. And if it works and we attain the kinds of efficiencies and qualities from this change that we think we can, as well as the cost savings, then analysis will be performed to see whether the concept can be applied to other things in the System. And if that fits, and if it makes sense to do it, I would assume the System would move in that direction. But I couldn't predict where that might fit. It's obviously important that we maintain the individual integrity of each district, so everything that's done needs to reflect that goal. We certainly don't want to do something that would fundamentally change the structure of the Federal Reserve System because of efficiencies we could gain in automation. I certainly don't think that what we're doing now has that potential, and I hope that nothing we do in the future would have that potential.

Region: On a more personal note, you were first vice president at the San Francisco Fed before assuming your current duties; why did you decide to take on this new responsibility?

Powell: That's a real difficult question to answer, and there's no single answer to it. Being a first vice president is a terrific job and a tremendous responsibility, and I personally think that the job in San Francisco is one of the best and most challenging in the System.

But I guess the enormity of what we're trying to do with automation consolidation, and what I perceive to be the tremendous benefits for the System if it's accomplished well, was more of a challenge than I could walk away from; although I continue to ask myself quite frequently why I decided to do it, as I think about the size of the undertaking. It's one of the most dramatic changes that the System has ever attempted to make. It's just something that I felt I needed to do, and that may sound like "motherhood and apple pie," but it's true. It's a very important effort, and I've been with the System for over 20 years and I have a lot of dedication to it. And as I thought about the skills that would be needed to make it work, I thought I could contribute something positive. So I'm here.

Region: Your career with the Fed has taken you from the heart of the Midwest in Kansas City, Mo., to the West in San Francisco and now east to Richmond. Do you have any particular geographical preference?

Powell: Well, I must not because I keep moving around. I grew up in the Midwest, not far from Kansas City, so I obviously am familiar with that part of the country, and I enjoyed living in Kansas City and working at the bank there.

I'm not sure there is any place that compares with San Francisco, quite honestly. The beauty, the topography, the cultural diversity, all of the things that are there—not to forget the great wineries and restaurants—it's just a terrific place to live. I fell in love with it the first time I went there and have thoroughly enjoyed my 12 years there.

I hate to leave, but I'm also excited about exploring Richmond. It's a much smaller community and certainly different culturally and environmentally and so forth. But there's a wonderful amount of history in Richmond, and the cultural change will be interesting, and the people at the Richmond Fed are just terrific. They've been very helpful. I'm excited about it and don't really think of the move as from West Coast to East Coast. I just plan to enjoy the move and look forward to Richmond, although I will greatly miss San Francisco.

Very honestly, all the traveling that comes with this position may make an East Coast location more convenient anyway. When you work in an organization like this, you very often find yourself making trips to the East Coast, and those cross-country trips from San Francisco can be very hard and tiring. So, in one sense, it will make my life a little easier to manage by living on the East Coast.

Region: As a final question, do you have any other topics you would like to address regarding consolidation?

Powell: One thing that's impressed me is the amount of commitment there is to consolidation.

We've completed the site selection process very quickly and with a great deal of support from all levels of the System, and I anticipate that continuing. I expect this to be a very successful process, and when we've finished we'll have improved how we do things and we'll be glad we stuck to the effort. I'm looking forward to it.

The speed at which this process has been moving is a compliment to the System. The site selection process was the first major decision to be made, and it was a real test of our new management structure and of our commitment to the idea. We passed that test with flying colors. It has gone extremely well and it's a tribute to everyone in the organization.

Region: Thank you, Mr. Powell.