Because SDI is large, its growth means it is consuming a large and sharply growing share of DOD's budget for R&D. In 1984, it consumed 4 percent of the funds. That is up to 9 percent in 1987. According to the administration's request, SDI will use 12 percent of all R&D funds in 1988 and 23 percent by 1992.

Moreover, SDI is consuming a growing share of a relatively constant R&D pie. Between now and 1992, DOD is requesting little real growth in the total R&D pie and, as a result, other programs will experience some real declines.

These budget trends could continue beyond 1992. Currently some technologies in SDI are in advanced development. The administration will consider going to the more expensive stage of full-scale development in the early 1990's. In addition, increased research spending in the 1990's on SDI's companion program, the Air Defense Initiative, could exacerbate effects on other R&D programs. I think the trends clearly suggest that under the administration's plans some R&D projects other than SDI face tough fiscal times. The Congress has to weigh the adverse effects on those projects against SDI's importance.

Now let me turn to some trends within the SDI budget. Most funds are apportioned within five broad program elements. Figure 1 on page six of my prepared statement summarizes them for you. Two of them fund weapons that would be used to destroy incoming enemy missiles. One pays for kinetic energy weapons, those that would destroy an incoming missile by hitting it with an object; that technology is more mature and would be emphasized in any deployment in the 1990's. Another funds directed energy weapons, those that would destroy enemy missiles with bursts from laser or particle beam weapons; that technology is much less mature.

In addition to the two program elements that deal with weapons, there are three others. One deals with sensors, another with the overall architecture and "brains" of the system, if you will, and the last with support technologies like space transportation.

In addition to these five program elements, total SDI funds include a small amount for headquarters and military construction and a larger amount that would be spent by the Department of Energy. I am not going to deal with those funds further in my testimony.

Dollars in the 1988 budget request are spread widely over these five broad categories. The largest share, 29 percent, goes to the program element for developing sensors. The smallest, 12 percent, to the program element to develop the overall architecture or brains of the system. The two program elements that pay for weapons systems to destroy enemy missiles would each get about 21 percent of the request.

Compared with previous 5-year plans, there are some shifts in priorities among these five program elements-shifts that tend to equalize their share. There is really no divergence among those broad categories.

One specific concern about trends within the SDI budget involves a degree of shifting of funds toward technologies that could be used in a near-term deployment. By that I mean a deployment sometime in the next decade or early in the next century as opposed to one

in a later period. Let me discuss those shifts briefly and then address their importance.

There has been a gradual increase in the share of funding for kinetic energy weapons. These could be made available more quickly than directed energy weapons. Of the total dollars spent on the two program elements that research and develop weapons, the portion spent on kinetic energy weapons has increased from 38 percent in 1984 to 46 percent in 1987, and it will be up to about 50 percent in the requests for 1988 and 1989.

When one examines funding within the five program elements in more detail, there are other signs of emphasis on developing technology that would be useful in a near-term deployment.

In general, within the program elements, funding is largest in absolute terms and growing at a faster rate for experiments dealing with integration and demonstration of technologies as compared with funding for the development of technology. So we are starting to do experiments as opposed to developing the technology or emphasizing that. The tables at the end of my statement give you some detail, and there are specific examples of these trends in my written statement.

What is the importance of these trends? SDI technologies are in varying stages of maturity, but the goals of SDI managers have always been to develop technologies sufficiently to support some full-scale development beginning in the early 1990's. To achieve that goal, they must spend relatively larger sums on more mature technologies and they must begin to emphasize experiments dealing with those technologies. The trends in funding noted seem quite consistent with that goal. On the other hand, these trends may mean less emphasis on infant technologies, and most analysts would agree that these infant technologies are also the most likely to eventually achieve an effective defense against ballistic missiles, especially because the Soviets will have more difficulty countering these approaches.

Hence, I think you face a dilemma. Does one emphasize more mature technologies now in order to be able to make a full-scale development decision in the 1990's, but risk shorting those higher payoff, higher risk technologies that are now less mature? Or does one risk delaying a decision on full-scale development in order to continue emphasis on the less mature technologies? That dilemma may soon become more pressing. According to General Abrahamson, the choice of whether or not to emphasize more mature technologies will have to be made more clearly if the Congress decides to place further restrictions on SDI's budget request.

Let me summarize my key points for you, Mr. Chairman. The SDI budget is very large compared with any other research project. SDI's budget is really more comparable to the R&D budget for a military service than the budget for a particular research project. Under the administration's request, SDI will also grow sharply as a percentage of total R&D funds to 23 percent by 1992. That will leave fewer dollars for other research projects.

In its 1988 request, SDI would continue to spend widely in a variety of technologies, but we do see signs of shifts toward more mature technologies that would be used in the near-term deployment. If budget limits continue to be imposed on SDI, the adminis

tration and the Congress may soon have to make a clear choice: either to slip SDI's timetable in order to hold down costs, while maintaining a balance of spending on mature and less-advanced technologies, or to emphasize those technologies that could lead to deployment in the 1990's and hold down costs by delaying higher risk, higher payoff technologies. I think this choice is one of the key decisions that you face regarding the SDI budget, particularly if you decide to place further limits on it.

Mr. Chairman, that completes my statement. Ms. Dombey and I would be glad to answer any questions you have.

Prepared Statement of Robert F. Hale

the

: Mr. Chairman, thank you for the opportunity to discuss the budget of the Strategic Defense Initiative (SDI). For fiscal year 1988, the Administration has requested $5.2 billion for SDI research in the Department of Defense (DoD). Over the next five years (1988-1992), request totals about $37 billion. (The total request for SDI research-$5.8 billion in 1988-includes about $600 million for work to be done by the Department of Energy, but I will focus on DoD funds.)

My testimony describes this SDI request and discusses some important trends in funding. Those trends suggest that SDI will consume a sharply growing share of all DoD research funds. There are also shifts in funds. within SDI that suggest growing emphasis on more mature technologies, though SDI continues substantial funding for many types of technologies.

It is beyond the scope of my testimony to judge the desirability of

these requests.

SDI has a far-reaching goal: to deter nuclear war by defending populations against a nuclear attack rather than by relying primarily on retaliation to provide deterrence. The desirability of. SDI's funding requests depends on an assessment of the importance of that goal and the likelihood of achieving it. It also requires examining the impact of SDI funding on other research programs for defense.

OVERALL SDI BUDGET TRENDS

The Administration is requesting $5.2 billion of DoD budget authority for SDI in 1988. The 1988 request makes SDI by far the largest single program within DoD's budget for research, development, test, and evaluation (often shortened to research and development or R&D). The next largest R&D funding request for which data are publicly available is for the small ICBM

at $2.2 billion. Excluding SDI, the three most expensive R&D programs in 1988 average $1.5 billion each. Indeed, SDT's 1988 budget request exceeds that of any single procurement program and rivals the R&D budget of at least one of the military services. The Army-the service with the smallest R&D budget is requesting $5.5 billion for R&D in 1988 compared with $5.2 billion for SDI.

Over the next five years, requested SDI funds for DoD would grow in real terms by an average of 14 percent a year to $9.8 billion in 1992 (see Table 1). Total five-year funding would amount to $37.1 billion. SDT's rapid growth is not unusual for a research and development program. Requested real growth is, however, sharply higher in 1988 (at 56 percent) than the average growth over the next five years. This request is presumably to make up for reductions in SDT's budget in the previous year (higher growth in the first year has been a characteristic of recent budget plans for SDI).

TABLE 1. TRENDS IN SDI RESEARCH BUDGETS (In billions of current dollars)

Excludes supplemental appropriation request of $0.8 billion. If this request were included, real growth in 1987 and 1988 would each be 38 percent.

Because SDI is large, its growth means it is consuming a large and

sharply growing share of DoD's budget for research and development. In