The Bank's quest for efficiency also was aided by a series of events that placed Atlanta in the somewhat surprising position of being the System leader in technological innovation, particularly electronic payments. This occurred just when such innovation started to pay big dividends in increased operating efficiency. The Atlanta Bank's technological ascent had begun inconspicuously with the formation of the Methods and Systems Committee in 1957. One of the first leaders of Methods and Systems was Brown Rawlings, who by the time he retired in 1982 would be hailed as "the father of the electronic payments system." In 1961 Rawlings, responsible for centralized systems planning for the Bank, contemplated a largely manual operation that was ripe for automation. Transaction volumes were rising rapidly. Without technological intervention the payment system soon would begin to choke on the deluge of paper. Vendors were rushing then to perfect machines that could process checks automatically. Moreover, the technology was forming that would allow the transactions themselves to become electronic and remove tons of clumsy paper from the system.
Rawlings was not a computer scientist or even an engineer. He was an agricuitural economist who worked for the U.S. Department of Agriculture before joining the Bank in 1947. But he saw the need for a technological transformation that would produce a modern payments system, and for the rest of his career he became a facilitator of that process. He recruited the engineers and computer scientists who could build such a system. Then he put them to work while he ran interference for them in the Bank and in Washington, protecting their freedom to experiment and find solutions.
The electronic payments project
One of Rawlings' most noteworthy alliances was with Professor Paul Han, a Georgia Tech electrical engineer who wanted to build a prototype electronic payments system and needed a bank, ideally a central bank, where it could be tested and deployed. Together Han and Rawlings turned Han's idea into a joint project. The Atlanta Fed would underwrite a four-part study by Georgia Tech on electronic payments. Banks in Georgia and Florida would provide the test market, and Rawlings would watch over it for the Fed and provide resources from the Bank at appropriate times.
Atlanta's position as technological pioneer was enhanced with the automation of the cash processing function.
The Board of Governors was eager to see the results and encouraged the Atlanta Bank to move ahead with the project. In May 1969 the Atlanta board authorized a $310,000 contract with Georgia Tech to carry out Phases I and II. The work was completed by August 1970. It constituted, Rawlings reported, the information needed to build an electronic payments system. The Bank was asked to authorize $456,000 for Phases III and IV, and Washington was ready to proceed. A second contract was signed in July 1970. When Dr. Han died suddenly in February 1971, Dr. Allen Lipis, a project team member on loan from First National City Bank in New York (now Citicorp), took over as director. By August 1971 the project had grown into the Committee on Paperless Entries (COPE), and the five largest Atlanta commercial banks were involved. They had formed a payments system steering committee and released eight of their top technical people to work full-time on the project. An attorney from the New York Fed and some marketing consultants also joined the project team. It wasn't the Manhattan Project, but it was turning into an electronic payments equivalent.
Expenses for the Georgia Tech work on Phase III alone grew to $421,000, which was $79,000 over the original contract amount. Kimbrel asked for continuing support, noting that the project had the potential to transform the payments mechanism of the nation, and the Bank came up with the extra money. Phase III was officially completed in March 1972.
The result was a ground-breaking electronic payments system. Atlanta's commercial banks agreed to foot most of the bill, and the Atlanta Fed agreed to operate the automated clearinghouse that would process the transactions. The goal was to replace the checks used for payrolls and bill payments with electronic entries. Thus Atlanta's COPE project and California's SCOPE (Special Committee on Paperless Entries) project begun in 1968 provided the first automated clearinghouses (ACHs) to process and clear electronic payments. The Atlanta ACH was operational by May 1973. Volumes were small at first, but the experimental batch processing system worked and became a pioneer of the U.S. electronic payments system and a prototype for ACHs in other Districts.
From Washington, Governor George Mitchell, a strong advocate of electronic payments, favored Atlanta as a test market for payments system experiments. Kimbrel's close ties to the bankers of the District assured bank cooperation, and Georgia Tech and the COPE project were marshaling the resources. Through the first three phases, the Board of Governors gave strong support to the project and watched the results with interest.
Phase IV began to take shape as an ambitious and expensive real-time point-of-sale system, and the Atlanta Fed, working with its Phase III partners, the major Atlanta commercial banks, was to operate the switching center. The project carried an estimated price tag of $4.2 million, of which the Atlanta Fed proposed to pay $1.5 million and the large Atlanta commercial banks would pick up the remainder. That, the Board of Governors decided, was sailing a bit too far into uncharted and commercial waters. It denied the Atlanta Bank's request to proceed. Without the Federal Reserve Bank the project collapsed, and the ambitious point-of-sale system being planned in 1974 still has not come about.
Atlanta continued to be a technological pioneer when cash processing operations were automated later in the 1970s. The System's first production model of the currency verification, counting, and sorting machinery manufactured by Recognition Equipment Incorporated was installed in the Bank's head office on March 6, 1978. These machines finally brought into the computer age the tasks of conducting a bill-by-bill verification of the amount of currency shipped to the Fed by depository institutions, detecting counterfeits, and identifying and shredding unfit notes. Where it once took 21 employees to process 760,000 bills in a day, 15 now could deal with more than 2 million bills.