NSF, DOD R&D slated for small increase

By Fred W. Weingarten
CRA Staff

With this year's budget and appropriations cycle still in shambles, the administration has moved on to next year's funding cycle with its release of the fiscal 1997 budget.

President Clinton is required by law to submit a budget at this time. So the release is not just a White House maneuver to show up a Congress that still has not passed all the 1996 appropriations bills. However, the administration used the opportunity to point out the irony. Of course, it is also an unwritten law that the congressional leadership will declare the budget "dead on arrival."

This ritual takes place no matter which party controls the White House, Senate or House. The president submits the budget; Congress reminds him that they control the purse. This year congressional rejection had a particularly authentic ring to it because of the exceptionally bitter history of the 1996 budget negotiations and the inevitable politics of an upcoming presidential election.

Although the budget as a whole faces a tough battle before it passes, it does contain useful clues to administration priorities for those concerned about research funding. After all, the appropriations bills for the two agencies of particular importance to computing research--the Defense Department and the National Science Foundation--passed last year, and both agencies were treated relatively well. Much to NSF's dismay, the larger appropriations package that contained its funding was vetoed. But the veto was aimed at the Environmental Protection Agency appropriations and had nothing to do with NSF.

The history of last year's congressional action suggests that the administration's R&D budget for these two agencies may not be far off the mark. Both the longer-term trends and the impact of the new Congress can be seen. For several years Defense R&D had been on a downward trend, which is no surprise. But the 1997 request of $38.3 billion is up slightly (.005%, or $172 million) from the 1996 estimated spending total. The push from Republicans to reverse the trend is also apparent, and the president's 1997 budget has reversed course in response.

NSF budget

For the past several years NSF's budget had been increasing--until the 1996 congressional action that, for the first time, resulted in a very small decrease from the previous year's actual spending total. That decrease occurred despite relatively benign treatment in Congress last year.

The realities of balancing the budget on the back of discretionary spending is no doubt what we are seeing. NSF's 1997 budget request is for $3.3 billion, up 4.6% from the 1996 spending estimate of $3.2 billion.

An important message contained in the budget is a sense of administration priorities. In NSF's request, computing research fared no better or worse than most other major scientific areas (see Table 1).

The detailed numbers for the Computer and Information Science and Engineering Directorate show some indication of priorities (see Table 2).

NSF's three research divisions are each slated to receive significant boosts at the expense of the two infrastructure divisions: Advanced Scientific Computing, and Networking and Communications Research. The Cross-Disciplinary Activities Program receives the largest increase in terms of percentage and increment. CDA increased 33%, from $24 million in 1996 (estimated) to $31.9 million in 1997. This increase is due mostly to a $6 million rise in funding for instrumentation, part of an NSF-wide program initiative.

Although it does not show up explicitly in Tables 1 and 2, the CISE budget contains a $4.5 million reduction in its share of the funding for the High-Performance Computing and Communications (HPCC) program. Overall, NSF's contribution is down by $1 million, although the total interagency budget is up slightly. It is unclear whether this drop in NSF support also reflects a reduction in administration enthusiasm for HPCC.

HPCC

The HPCC program, a cross-agency initiative of the National Science and Technology Council, has been given a small increase of 3% overall. More interesting is the slight shift in funding priorities. As noted above, NSF participation is down by 4%, while Defense (DARPA) spending is up by 7%. The Transportation Department is nearly doubling its contribution, from $23 million to $43 million (see Table 3 for a breakdown by agency).

The DARPA and NSF numbers, although not the largest increases or decreases in terms of percentage, are particularly important because the two agencies together represent more than 60% of the total program funding and represent the heart of the basic research funding.

Also of interest, the Education Department increased its participation by 50%. This increase in the HPCC line, a reflection of the high priority the administration has placed on educational technology, illustrates the department's opinion that fundamental advances in computing technology will be needed to achieve real educational benefits from information technology.

As the budget is dissected and analyzed in the appropriations process, more details will emerge. Submitting the budget is only the first step in a long process that, as we have seen this year, can turn into an agonizing, never-ending story. But the basic message is hopeful for the short term:

• Despite enormous budget pressures, research is still in relatively good political standing.
• Despite fears to the contrary, computing research is still a strongly supported part of NSF's program.

However, only an extraordinary optimist would take either message for granted over the long term. Long-term prospects still look grim.