System Development Foundation (Palo Alto, Calif.)

SDF was created in 1969 as a non-profit funding organization during the transition of the System Development Corporation (SDC) to for-profit status. The SDF Board of Trustees gave its initial grants to non-profit organizations, particularly Cal Tech, U.C. Berkeley, The Rand Corporation, and Stanford University. They backed research primarily in information sciences and the human-machine interface, the areas of expertise of SDC. In the first year SDF funded over $11 million in research; in the next five years it committed itself to $25 million. Recognizing it would terminate its operation in June 1988, funding was stopped for new projects in 1984. In the remaining years they continued to support work already underway.

Biography

The System Development Foundation (SDF) got its start in the 1950s as a not-for-profit corporation researching and developing computer software in the information sciences. To clarify its rather complex origins, which go back to The RAND Corporation of Santa Monica, California, a bit of background information is required.

ORIGIN OF SDF In the 1950s, The RAND Corporation, a not-for-profit United States Government think tank, experimented with a new system for training air-traffic controllers. Rand developed computer simulations for all the necessary instructional steps and test situations and demonstrated that people newly trained on this system could outperform experienced people trained in the program then used by the United States Air Force. The Air Force funded further development, and RAND created a new System Development Division to handle this and similar projects.

Soon the new division was growing rapidly, and its work was beginning to diverge from the think tank's stated mission. As a result, in November 1956 it was separated out and incorporated as the System Development Corporation (SDC), an independent not-for-profit organization. It would maintain its close ties with RAND Corporation.

At the time, there were so few companies experienced in the uses of computers or in developing computer software that the Air Force did not require open competition for the contracts it awarded SDC. During a decade of such preferential treatment, up to 1966, SDC's operations grew significantly as the corporation branched into other software markets, such as the computerization of Medicare's administration and computerized typesetting.

In 1966, there was a change in Air Force policy that profoundly affected SDC and eventually resulted in the formation of SDF: henceforth, open bidding would be required on its software-development contracts. Essentially, SDC was thrown into the commercial market. Within three years, SDC would request permission to become a profit-making organization.

The records in the Stanford collection reflect at least two different ways of looking at this change. One view, held by the Internal Revenue Service (IRS), is that SDC had been operating progressively less like a not-for-profit organization even before the Air Force decision. The second view is the one held by SDC--that, when faced with the Air Force's new policy, the executives and Trustees of not-for-profit SDC decided that the corporation had profit-making potential. They then began planning how to reorient its legal status. By 1969 their plans were complete and they proposed that SDC change from a not-for-profit to a profit-making corporation. In their opinion, SDC's competition for contracts during the previous three years did not represent an earlier shift towards profit-making status.

The Air Force, the IRS, and the California State Attorney General all had reservations about this proposed change of status. SDC had been tax-exempt throughout its lifetime. Moreover, for all these years it had received contracts from tax-supported institutions without having to bid for them.

Negotiations eventually yielded a compromise. The not-for-profit corporation would retain its tax-exempt status but take a new name: The System Development Foundation (SDF). This Foundation would own most of a new, for-profit corporation which would now be known as The System Development Corporation (sometimes referred to as "new SDC" or "SDC #2"). The Board of Trustees of the original, not-for-profit SDC would become the Trustees of SDF, and a new Board of Directors would quickly be chosen for profit-making SDC.

In return for a controlling share of the for-profit SDC, SDF transferred to the new corporation all of the actual facilities, operations and contracts of not-for-profit SDC. In essence, one organization split into two, a profit-making corporation receiving the established business; and a not-for-profit foundation receiving a large share of the corporation's stock. This, however, is not how the Trustees and Directors of SDF and SDC--or the various concerned branches of the Government--saw it. Because of the tax-exemption, they all considered SDF to be the same legal entity as the old not-for-profit corporation, and for-profit SDC an entirely new corporation. The Foundation's corporate seal, for example, shows that SDF was incorporated in 1956, completely ignoring the fact that it was known by another name at the time.

Because the original SDC had been tax-exempt for more than ten years and had been treated preferentially by government agencies during that time, the Government was concerned that an established business would be appearing whose start-up funds had in effect come from the U.S. Treasury. The Trustees also wished to avoid this perception and any possible back-taxes or penalties such a perception would provoke. Thus, a further part of the agreement was that SDF would pay the Treasury half the value of for-profit SDC. SDC was valued at eight million dollars at the time of the reorganization. SDF would therefore pay four million dollars to the Treasury as a "voluntary contribution": two million dollars to be paid immediately with funds borrowed from SDC; and two million more when SDF sold its stock in SDC. (SDC's Board of Directors and SDF's Board of Trustees intended that SDC stock would be offered on the open market.) SDF's role would be to distribute to the American taxpayers the money put into SDC by Government contracts. The money to be distributed--like the second $2 million in direct payments to the Treasury--would come from the eventual stock sale. With the proceeds of this sale, SDF would make grants to tax-exempt institutions sponsoring research in the fields in which the original SDC had been active.

The agreement was eventually approved, but SDC's competition was not pleased. The recently formed Association of Independent Software Companies lobbied intensively to block the conversion to for-profit status, claiming among other things unfair competition stemming from SDC's growth as a sole-source government contractor. SDC eventually agreed to make available its major findings, research results and computer programs developed during its non-profit period. This helped alleviate some concerns, but not all as indicated by the headline which Business Week used to announce SDC's reorganization, "Software Giant Goes Commercial."1

The change in status occurred in July, 1969. Once SDC's Trustees had selected a new Board of Directors, they relinquished direct responsibility for SDC and assumed direction of SDF's affairs. SDF's original Board of Trustees in 1970 consisted of Arnold O. Beckman, Edwin E. Huddleson, Jr., Augustus B. Kinzel, Lloyd N. Morrisett, Lt. General Donald L. Putt, Dr. Horton Guyford Stever, Dr. Ralph W. Tyler, and Bethuel M. Webster. With the exception of a Trustee who had resigned in 1968, this was the same group which had initiated the change to for-profit status in 1967. In addition, Huddleson had been one of the original founders of the System Development Corporation in 1956. Following the reorganization, the SDF Trustees and SDC's Board of Directors held periodic joint meetings. The Directors made all business decisions but were required to keep the Trustees, as major shareholder, aware of all their proceedings.

Three separate plans to make stock available to the public were frustrated by circumstances beyond the control of those involved. The first offering was scheduled for December 19, 1972, the day after President Nixon announced the resumption of the bombing of Hanoi. This event sent the stock market into a downslide and the offering was cancelled. Attempts at public stock offerings were made again in 1973 and 1976, but each time unfavorable market conditions resulted in the cancellation of the offering.2

The Treasury Department became uneasy with the situation because no grants seemed forthcoming, and the second two million dollars was likewise delayed. The Internal Revenue Service again questioned whether non-profit SDC ought to have been accorded tax-exempt status in the last few years before 1969. The IRS also was suspicious about the nature of the voluntary donation to the U.S. Treasury. To prove their intent to give away the money, the SDF Trustees made a series of preliminary grants. In 1974, the following four institutions each received 50,000 shares of SDC stock to support their research programs: the California Institute of Technology, The RAND Corporation, Stanford University, and the University of California at Berkeley. There were restrictions on the gift: the shares were not registered with the Government to be sold openly, and shares paid virtually no dividends. The gifts were accepted, however, and each institution handled the gift according to its policies. Stanford made an internal transfer and freed funds for immediate use; the other institutions waited until a tender offer was made by SDC and by the outside corporation which would eventually purchase SDC for the stock they held.

An additional restriction on the Foundation was that its funds must be dispersed within a certain time frame. By the end of the 1970s, SDF's funds were still tied up in SDC stock. The Trustees decided that they could no longer accept the responsibility involved in waiting for a successful public offering of stock. They located an investment group, Lehman Brothers, willing to purchase their shares and take over the financial risk. SDF would then have the cash to begin an active grant program. However, the California State Attorney General refused to approve the agreement. The Attorney General pointed out that it had been Lehman Brothers, as independent consultants to SDC in the 1960s, who recommended that SDC go commercial. When this deal fell through, SDF sought another purchaser. In December of 1980, a controlling interest in SDC was sold to the Burroughs Corporation, and SDC merged with this larger organization.

The history of SDC from its inception until the Burroughs merger is detailed in Claude Baum's above-referenced corporate history, The System Builders: The Story of SDC (available in Special Collections Rare Books Reference and in Green Library stacks at HD9696.C64 S83 1981). Baum's primary focus is on the Corporation during the years 1969-1980, not on the Foundation.

DEVELOPING THE PROGRAM Once SDC was sold and funds became available, the Trustees of SDF faced the task of developing an active grant-making program. As Director of Programs they appointed Charles S. Smith, formerly with the Department of Energy in Washington DC. He and a small staff began work in 1981.

Shortly after SDC made its grants of stock in 1974, it began to receive requests for money. Requests were put on file because the Foundation was not yet active. The publicity surrounding the sale of SDC generated still more interest. Early requests were politely acknowledged and put on file until a program could be created. It may have been hoped that these files could serve as a resource in developing the program itself. In 1980, Ralph W. Tyler, President of the Board of Trustees until SDF's close in 1988, began asking other Trustees and contacts across the country to suggest people who might benefit from--or help develop--the Foundation's program. Suggestions received by Tyler and information gathered by Smith and his staff about SDC's former fields of concentration and the people involved formed the basis of the Foundation's early grants.

SDF intended to fund research in those areas in which the old not-for-profit SDC had worked. Smith's first goal was to inform the Trustees of current research and funding patterns in these fields. To this end, Smith and his staff arranged a series of seminars in fields in which SDF would likely fund research. The seminars were also intended to survey existing funding patterns, because the Trustees wanted to complement available funding sources rather than to compete with them. Specific titles of the seminars are mentioned in SDF's 1982 Annual Report.

Presentations by invited experts gave the Board the necessary background for formulating a set of funding guidelines. These guidelines, issued in December 1981, declared SDF's intention to support research in principles of information science, computer representation or modeling in biological and machine information processing (as in neurobiology, cognitive sciences, non-Von Neumann computer architecture, and robotics), principles of human-machine interface, and the interface between the computer and artistic endeavor. These were eventually formalized into eight areas that would remain constant for the life of the Foundation: principles of information science, computational linguistics and speech, symbolic mathematics, robotics, non-Von Neumann computation, neurosciences, human-machine interface, and computer music.

Charles Smith was now authorized to solicit proposals actively. Likely proposals were presented for action to the Board as a part of the materials to review for their meetings. During this time, Smith also contacted experts in the various areas of interest to SDF. These reviewers and consultants proved very helpful in developing SDF's policies, and the policy of confidential review by outside experts that developed from this consultation policy would become a permanent feature of the Foundation.

SDF issued its first annual report on June 30, 1982. It noted the Trustee seminars, outlined the areas of interest as formalized in December 1981, and contained recommendations on how to apply for funding. It was hoped that this report would reduce the number of out-of-scope proposals received, especially the charitable requests, as well as publicize SDF's efforts more widely among those who would benefit from its funding policies.

There soon developed a regular pattern of proposal and response. Initial contact could be made either by SDF or by an interested researcher, but most often interactions began with a recommendation by someone already familiar with or associated with the Foundation. Researchers were asked to submit a short proposal, along with related reprints to serve as background material and to establish the credentials of the investigating team. If such a proposal seemed to be in line with the SDF program, it was sent out to independent reviewers. Sometimes Charles Smith or a reviewer would visit the researcher's laboratory to get a clearer sense of what was hoped for in the proposed project. Smith often worked with researchers who had promising proposals, helping them to clarify their research questions and to make a clear statement of their goals in the funding request.

Quality of research was not the sole criterion upon which the Trustees based their funding decisions. For example, SDF wished to commit its resources to important projects that seemed unlikely to receive money from other institutions. Smith sometimes helped researchers revise their proposals for submission to the National Science Foundation or the Department of Defense. If a researcher seemed already occupied with a multitude of projects or duties unrelated to the proposed research, SDF was hesitant to fund yet another project, considering that it might not receive enough attention. On the other hand, at times it was obvious that special funding of a researcher's project would alleviate pressures of other university duties; this would have been an additional reason to award a grant.

When the staff had gathered the outside reviews of a proposal, it was brought before the Board of Trustees. The Trustees met quarterly, and their agenda materials contained information on each proposal to be considered. For the first year of operations, the Trustees insisted upon receiving copies of all documents related to the proposals under consideration. By December of 1982, however, it was clear that this would not be feasible. The volume of paper was simply too great, and the material was too technical for the Trustees to absorb. The Board decided to rely on the recommendations of the reviewers, who had been chosen because they were experts in their fields. The SDF staff wrote summary sheets for the Trustees, collating the information gathered about each proposal.

On the basis of these staff summaries and the discussions during Board meetings, the Trustees decided whether to make a grant or not. Sometimes, proposals would be partially or conditionally funded, or the principal investigator on a particularly promising project would be asked to explain more clearly how the project fit in with SDF's program. On rare occasions, a small grant would be made to cover development costs for a larger proposal.

From the early 1980s onward, SDF had intended to operate for a limited time. Its Trustees believed this would ensure that their funds would have a larger impact than if diluted across a longer period. In 1984, the Foundation reached a point where half of its resources had been committed. The Trustees decided to cease encouraging new proposals and to concentrate on previously funded projects or closely related ones. To this end, they reorganized the Foundation somewhat. Charles Smith resigned as Director of Programs, agreeing to remain affiliated with SDF as an independent consultant. The Director's job description was altered; and in August of 1984, Carl M. York became Director of Program Administration.

Before York arrived, SDF's clerical staff cleared the files of proposals that had been superseded or seemingly forgotten by their researchers, as well as those that were inappropriate to the Foundation's new focus.

MAJOR GRANTS AND SPONSORED PROJECTS SDF funded research across a wide area; for a thorough discussion, the researcher is referred to the Foundation's final annual report, appended to this collection guide.

In the mid-1980s, the staff of SDF began investigating the possibility of obtaining a reduced price for computer equipment for its grantees. Xerox Corporation's then-new Dandetiger computers became the focus. Eventually several researchers opted to use some of their SDF funds for this purchase.

The 1984 Annual Report notes SDF's interest in holding benchmark symposia in its funding areas. The 1986 Annual Report announces dates for three symposia in neuroscience, robotics and computer music and for four workshops, all in computational linguistics. Original hopes to hold symposia in all eight areas of concentration were abandoned after consultation with researchers in the fields. One researcher in computational linguistics, for example, suggested that it might be premature to hold a retrospective symposium in such a new field. He felt that the effort and money put into summary papers would be better spent on pushing forward--examining research goals and exchanging ideas on current research. The four computational linguistics workshops were designed to bring together researchers in that field so that they might exchange information and learn about one another's progress.

WINDING UP THE FOUNDATION As early as August 1984, when York arrived, the staff of SDF knew that the Foundation would close about 1988, but the exact date was not firm until 1986. In that year, the closing date was finalized for June 30, 1988. Staff members were asked to work full time through 1987. Anyone willing to work up to or beyond June 30, 1988, would be eligible for severance pay. Some members agreed to work beyond that time as consultants, to assure the proper winding up of SDF's responsibilities.

SDF was concerned about grants it had issued whose terms went beyond, or might be extended beyond, that time. Many ideas were suggested to deal with this eventuality; the final decision harked back to SDC's original articles and bylaws of 1956. These dictated that, should the not-for-profit organization finish business with a financial surplus, it should give that surplus back to RAND or to a similarly tax-exempt research institution. SDF arranged to make a set of closing grants to The RAND Corporation to ensure proper administration of continuing research projects and proper fulfillment of SDF's legal obligations. RAND received approximately $7 million and assumed legal responsibility "up to but not exceeding the amount of the grant."

Toward the end of SDF's lifetime, its trustees began to explore possible ways of measuring the Foundation's impact on the areas funded. The Board chose to sponsor a series of symposia, to place SDF's records in an archive, and to fund a history of the Foundation. The Stanford University Libraries' Department of Special Collections was chosen to house the records, and a grant was made to process the collection. An additional grant was made to the University Libraries for funding documentation and research projects to evaluate SDF's impact on the evolution of contemporary science and technology.

Principal investigators for grants active after SDF's final dissolution in June 1988 were requested to send their expenditures reports to RAND and to send research reports to the Stanford University Libraries for addition to the Records of the System Development Foundation.