Policy 101: Science Policy and the Computing Societies

By Fred W. Weingarten

Introduction

My current role with CRA as Director of Public Policy is drawing to an end. This is a good opportunity to look back over eight years of representing computing research in this town, and assess over the course of two or three articles in general where the field now stands in the science policy debate. In doing so, I’m not going to spend a lot of time in the past. Opportunity always lies in the other direction, toward the future, and I think that computing research will be faced with enormous opportunities over the next few years. In these articles, I’m going to briefly address three basic issues.

In this piece, I’m going to describe briefly the nature and structure of the science policy debate; in January, I’ll discuss the decision-making process and how to influence it. In a final piece in March, I’ll explore what the computing research community, itself, should consider doing, especially in light of recent developments, particularly the opportunities presented by the Interim Report of the President’s Information Technology Advisory Committee (PITAC) (see “PITAC’s Interim Report: Expeditions Of An IT Kind,” page 1). March is when the two-year political cycle really starts; it is also when the computing research community should start any campaign to improve the state of affairs of computing research funding along the lines of the recommendations of the report.

Computing Research at the Table

At the Snowbird ski resort in Utah in the summer of 1988, computer science and computer engineering department chairs decided at their biennial meeting to reconstitute what was then known as the Computer Science Board as a formal organization. The organization was to be called the Computing Research Association and its offices would be located in Washington, D.C. The choice of a Washington address was deliberate. The principal purpose of this organizational change, as some explained later, was to give computing research “a seat at the table.” The Snowbird group felt that, for too long, decisions regarding the nature and evolution of their field were being made by people who were, at best, ignorant of its particular needs or, at worst, hostile to its interests.

Coming into existence only in the early 1960’s, the fields of computer science and computer engineering are much the junior to the traditional fields of research in the natural sciences, whose origins date back several centuries or even millennia. Over the preceding decades, computing researchers had seen federal research policy and funding priorities set predominantly by representatives from these older fields, most of whom were much too old to have been exposed to computers or computer science in their graduate training or research.

It is commonly said that the heavy representation of physicists in post-war science policy circles was a reward for the important contributions they made during the war. My guess is that their willingness to engage in politics came about because they realized, first, that the well-being of basic research in physics had become inextricably linked with federal government support — if for no other reason, because of the rapidly escalating price for experimental facilities. Second, some of these scientists, confronting the awful (and awesome) results of their work, felt that an enormous sense of social and political responsibility had been laid on them.

But now, computing and digital communications, technologies that had their origins in that same war and that grew in the post-war environment, are assuming major social and economic significance. At that 1988 Snowbird meeting, attendees declared that computing research had come of age, and it was time for the field to assert a voice in the policy debate. CRA in its current form was created to carry out that mission.

Having Something to Say

Ah, but how to do so? It is one thing to have a seat at the table. One simply buys a chair and hires someone to sit in it. That was me. It is quite another to have something useful to say and to influence the debate. While I have rarely been accused of lacking something to say, and I did the best I could, it has taken the field some time to develop a coherent message that everyone could agree on. Looking back, it seems symbolic to me that shortly after I arrived at CRA, the field erupted in a fight over the NRC report, Computing the Future, A broader Agenda for Computer Science and Engineering, 1992.

The specifics of that fight are not important, but it illustrated the difficulty computing research faced in getting its story together and, even more importantly, in getting it commonly agreed on within the community.

We are being asked with great urgency by the political community to participate in policy making. It is no longer necessary for us to convince political leaders that computing research is important to the nation; that is broadly accepted as a proposition in all the right places. But in politics, good will does not translate automatically into good policy, or even tangible support. Now, we have to answer a broad range of questions from policy makers and politicians whose bottom line is: “What should we do about it?” And, of course, if computing researchers don’t provide a convincing answer, others are standing by, ready to do so for us.

Answering that question is not easy. Science policy is a complex set of issues, and a fifty-year history of debate and many established precedents –– customary ways of thinking –– underlie the present system. It is not easy to say something new, even if we do have a sense that the present system is not serving computing research as well as it might and that something new needs to be said.

So, the computing research community has to undertake two tasks to prepare itself to assert an effective voice on science policy: develop its message, and organize to deliver it. The organization I will leave for the next articles in this series. Here, I want to concentrate on the structure of policy itself.

The Bottom Line of Science Policy

It is often said (even by me) that the bottom line of science policy is the dollar sign, since the principal focus is on government funding of research. But, in fact, the underlying structure of science policy is much more complex. Federal funds come by way of intricate, political decision-making — a careful dance among science agencies, the administration, the Congress, and the research community. This process assures that any decision to spend federal money on R&D, even for fundamental research, comes with strings attached.

Some of those strings:

Purpose: Whether we have a surplus or deficit in the budget, many more hands are out than there are funds available. All appropriations have to be justified and tied, in some way, to a public purpose. Pure philanthropy is not a particularly persuasive argument. That means that support of research has to be tied to national objectives: economic growth, national security, public health, national pride, and so on. Linking the two end points — politically acceptable purposes and the need to fund undirected fundamental research — can be a difficult, even a distasteful task, one for some in the scientific community, but in a democracy, it is absolutely critical to getting public money spent on science.

Priorities: Here is where the toughest fights will occur. Congress insists, and so far the scientific community resists the demand, that priorities be set. But priorities are and always have been set on science funding. When Congress decides to spend $10 billion on a superconducting supercollider, it’s setting priorities; and when it changes its mind, that, also, is priority-setting. Similarly the choice to spend $7 billion on basic medical research and $500 million on basic computing research clearly reflects, rightly or wrongly, current political priorities. So, the issue is not to set or not to set priorities; the issue is whether the scientific community will participate in the process. Many senior science policy people still argue against setting priorities, saying that science needs to be supported across the board, and that we need a broad spectrum of healthy research activity. Both ideas seem correct to me, not contradictory, but, then, I’ve been in Washington too long.

Processes: Here comes the bureaucracy! But, the fact is, to spend billions of dollars requires organization and process, and organization and processes can make a big difference on the community. There is a wide range of models already within the government, from NSF’s model of peer-reviewed individual investigator grants to support of massive federal contract labs like Los Alamos. Suppose one were to argue for doubling funding for basic computing research. Who should be the lead agency? How should it be spent? Who will decide what to spend it on and how will that decision be made? These decisions are sometimes made in a bureaucratic process that resembles trench warfare and that often takes place below the threshold of attention for the affected communities. But, the choices do have consequences, sometimes enormous ones. Again, since decisions on process will be made one way or the other, shouldn’t the community be part of the debate?

Proscriptions (Strings): Finally, we need to keep in mind that political decisions are not always rational (those of you in shock may close your eyes while reading this paragraph). Even straightforward funding or authorization bills will often come attached with odd and often intrusive little strings. Funds will be earmarked for various purposes or institutions, restrictions will be placed on the expenditures or on use of federal funds. Thankfully, much of this stuff goes away, but a lot of time is spent on it each year by the Washington science policy community, and sometimes it doesn’t go away.

So, science policy has these two interlocking and mutually dependent aspects:

• Its overall purpose, which is to keep the federal science programs well funded.
• All of the corollary but critically important considerations raised above (each of which raises a far more complicated and richer set of questions than I have given justice here).

Next month, I’ll discuss the political environment in which these policy decisions are made. Who are the players and what games do they play?