he next President, whoever he is, will find START III placed prominently on his agenda. Republicans and Democrats alike have complained that US and Russian nuclear arsenals remain too large, even after years of reductions, and they will demand more cuts.

On the surface, in fact, it appears the only remaining issue is whether START III's proposed reductions are of the proper size.

Appearances, in this case, are deceiving. Debate over the final warhead number masks an equally significant matter-the fact that START III will force major changes in the USAF bomber fleet, affecting a major element of US conventional power.

The START I treaty has limited US and Russian strategic warhead totals to 6,000. START II, which has been ratified but has yet to enter into force, would lower the number to 3,500, and the United States is already gradually reducing forces to that level. START III, outlined in 1997 but awaiting detailed negotiation, proposes a ceiling of 2,000 to 2,500 warheads.

The START III number will force harsh trade-offs in weapon systems. Even though bombers are prized for both their conventional and nuclear capabilities, the US may have no choice but to relinquish B-52H capabilities to meet START III limitations.

Secretary of Defense William S. Cohen warned in a May 28 NBC News "Meet the Press" broadcast: "As you get smaller, you may have to give up some of your bomber force."

That action, he added, "takes away from your conventional capability--the kind of capability we used in Kosovo."

Moreover, the bomber fleet will require significant modifications, and the US will have to cut the weapons deliverable by B-2s and B-52s. These changes are coming, but the US hopes to minimize the impact with some form of "counting-rule" relief through negotiations with Russia in finalizing START III.

Trials of the Triad

The problem stems from the US desire to preserve the basic composition of its nuclear deterrent. The US puts priority on maintaining a triad of land-, air-, and sea-based weapons.

The triad currently comprises silo-based missiles (500 Minuteman IIIs and 50 Peacekeepers); Sea-Launched Ballistic Missiles (432 Trident II D-5s); and bombers (21 B-2s and 94 B-52Hs).

The Pentagon will implement several force structure changes to reach START II levels by 2007. Plans call for Washington to:

• Scrap the 50 Peacekeeper ICBMs and keep only the 500 single-warhead Minuteman IIIs (500 warheads).
• Cut the number of B-52s to 75--perhaps with 43 carrying eight cruise missiles and 32 carrying 20 weapons (984 warheads).
• Reduce from 432 to 336 the number of Trident II submarine-based missiles at current force loadings (1,680 warheads).
• Maintain all 21 B-2 bombers, each loaded with 16 weapons (336 warheads).

These actions, taken together, will reduce American warhead totals from START I's 6,000 to 3,500.

That's the easy part. The next step--START III-brings serious force structure problems. While there are many ways to build a START III deterrent, all put the squeeze on the bomber force.

In START III planning, some forces are favored more than others. The Navy Trident II missile force, deployed on Ohio-class submarines, seems to be the most prized leg of the triad, given the submarine's stealthiness and relative invulnerability to attack.

Notional START III inventories typically show submarines accounting for more than two-thirds of total warheads under START III.

The ICBM force, slashed dramatically in recent years, also seems reasonably secure, if only because further shrinkage likely would make the land-based leg uncomfortably small. Also, ICBM cuts simply wouldn't put many warheads on the scrap heap.

One official said DoD doesn't rule out going down to 350 or so ICBMs, but cuts below that level would be problematic. "If you had one wing [of about 150 missiles] maybe you should think about getting out of that leg of the triad," he said.

By default, then, bombers have become the most prominent targets for START III cuts.

The threat does not affect all bombers. With only 21 aircraft, the fleet of B-2 stealth systems will surely be spared cuts or transformation to non-nuclear status. All B-1B bombers already have been shifted to conventional-only use.

All that remains for change, then, is the Air Force's fleet of 94 venerable B-52Hs.

Under START III, the Pentagon won't have enough warheads to distribute across all the delivery systems it wants to keep in the triad, and the BUFF looks like the odd man out.

Pentagon officials think the ICBM force's total of 500 warheads would remain unchanged. The US would "download" each Trident II missile from five warheads to four, shaving total SLBM warheads from 1,680 to 1,344. B-2s would be recalibrated, with each bomber modified to carry eight weapons, for a maximum of 168.

That Was Quick

At that point, the Pentagon will have used up 2,000 of a maximum 2,500 warheads. It will then have, at most, 500 warheads with which to equip its large B-52 force. (Obviously, the problem would be worse if the final number drops to 2,000 or fewer.)

Defense officials note that, even if the Air Force keeps only 75 B-52s, the service would have to drastically "downscope" each BUFF's weapon "loadout" to meet the 2,500-warhead limit. Today, B-52s can carry up to 20 nuclear weapons.

A Pentagon official identified one possibility: Keep 75 BUFFs (plus one START-exempt B-52 for test purposes) and declare them operative with only six weapons apiece.

"That would give you 450 warheads and all the B-52s could be available for nuclear and conventional missions," he said.

This move has a catch, however. "If you declare B-52s at six [weapons]," said the Pentagon official, "how does that impact conventional capability? There's a little monkey wrench in the process."

By that, he meant that the Conventional Air Launched Cruise Missile is a "counted" system under START; it is externally indistinguishable from the nuclear missile version. Notionally, then, each B-52H would be able to carry only six CALCMs, far fewer than it can carry today. B-52s now can carry 12 external and eight internal Air Launched Cruise Missiles or CALCMs.

The official said the US might try to "work something with the Russians that allows you to distinguish between the ALCM and the CALCM," but past efforts along those lines have failed.

The B-52 in recent years has been heavily tasked for conventional missions. The Air Force's 1999 bomber roadmap says the service will use it primarily as a cruise missile carrier until 2037, meaning that retaining the B-52's ability to launch missiles remains a high Pentagon priority.

Another possibility now being considered: Reduce the nuclear B-52 fleet to 61 and declare each bomber to be armed with eight cruise missiles, for a total of 488 accountable warheads.

Under this proposal, the other 14 B-52 bombers would be used exclusively as conventional weapon carriers.

Once again, however, there's a catch. A nation can maintain a "split fleet" of bombers of nuclear and conventional types, but different types can't be deployed at the same base.

Moreover, the US would have to prove that conventional B-52s are incapable of carrying nuclear weapons-not an easy task, given the similarity between the ALCM and CALCM.

In the words of a Defense Department official, "It will require a significant amount of modification. The details of how you'd have to change it haven't been worked out, but, clearly, you'd have to remove [mountings] so you can prove to the Russians that this thing's not capable of carrying the ALCM anymore."

That creates a new problem. Eliminating a B-52's ability to launch an ALCM would also remove its ability to fire the CALCM, a weapon used in many military operations in recent years and which no other USAF aircraft is configured to carry.

The simplest way to preserve the B-52's capabilities would be to obtain changes in counting rules in talks with Russia.

Ryan Seeks Relief

USAF's Chief of Staff, Gen. Michael E. Ryan, raised the matter in a May 23 Senate Armed Services Committee hearing on the nation's strategic nuclear forces.

"We need either counting-rule relief or reattribution as the numbers come down," Ryan told the panel. "That would be part of the negotiations as we went to that level."

Another US official noted that, under START II rules, the United States can "reorient" up to 100 heavy bombers from nuclear to conventional missions.

"One of the things we could do under START III is negotiate the ability to reorient" more B-52s, he said. This should be done, he said, because Pentagon civilian policy-makers "agree with Mike Ryan that B-52s have an important conventional mission, in addition to a nuclear one."

Another possibility for counting-rule relief: Focus more tightly on weapon systems that actually can cause destruction at any given time.

Officials note that B-52s undergoing extensive depot-level maintenance cannot launch nuclear weapons. Similarly, two of the Navy's Ohio-class strategic submarines are normally in overhaul and not usable for nuclear operations.

"So," the official said, "let's focus on actual shooters" and not focus on irrelevant weapons when determining START III counting rules.

Washington could greatly simplify the process by eliminating the B-52's role as a nuclear platform altogether and converting it to conventional missions. DoD officials are wary of making changes that would shake the stability of the triad, however.

When asked about the wisdom of eliminating nuclear capability, a Pentagon official said, "Those planes have some special capabilities that aren't available from any other leg of the triad."

For the Defense Department, the worst-case situation seems to be one in which START III cuts warheads below 2,000. Pentagon planners say, in their "what-if" scenarios, the arsenal begins to undergo strain when warhead numbers drop below 2,500.

For that reason, START III's declared upper limit of 2,500 warheads often is viewed as the level that will finally emerge in negotiations with the Russians.

A top DoD expert noted, however, that the number could just as easily be 2,000, which would mean "you've got to get rid of 500 weapons." That number corresponds to five Trident II-equipped submarines or the entire fleet of nuclear B-52s, the official explained.

"As you get to 2,000, it really starts squeezing the triad," said another Pentagon official. At that point, "you are clearly down to no more than two wings of Minuteman IIIs," which represents about 350 missiles compared to the 500 planned, and "you certainly don't deploy 14 subs. You may be down to 12."

"The bottom line is, it will require difficult decisions," said another DoD official. "Once you get below 2,500 [warheads] the decisions start to become very painful."

This is where counting-rule relief becomes critical. With some changes in how bombers are counted, said a DoD planner, "we could do a mix of those" while still preserving an acceptable nuclear and conventional force structure.

Out of Business?

The Defense Department is only too aware that cuts that go below 2,000 warheads are attractive to many in Congress and the Washington arms control community.

"You start thinking about 1,500 or 1,000 [warheads]," said an official, "and ... what do you do? You really are not, at that level, in the triad business anymore."

The question of what actually constitutes the proper number of nuclear warheads is highly contentious.

Sen. James Inhofe (R-Okla.), a conservative lawmaker who sits on the Senate Armed Services Committee, feels no reductions should be made at this time at all. Inhofe, through a spokesman, declared, "Moving to START III before you even get to START II doesn't make a whole lot of sense. ... Cuts should not be made before a full review, something a new President should be permitted to do."

An opposite view comes from Sen. Robert Kerrey (D-Neb.), whose state includes US Strategic Command headquarters at Offutt AFB. He calls for large and rapid reductions.

"Given our conventional and intelligence capabilities," explained Kerrey, "I am confident we can deter any aggressor with less than 6,000, or 3,500, or even 2,000 warheads. It is time we begin the process to come up with a realistic estimate of our deterrence needs."

Sen. Kent Conrad (D-N.D.) takes a position midway between Inhofe and Kerrey. He advocates reducing force levels to 2,500 but not lower.

In an April speech, Conrad said, "I personally would not be in favor of going down to 1,500 warheads. I think it's destabilizing. I think it absolutely shreds the triad. The triad is proven. It's worked. We shouldn't give it up."

Adm. Richard W. Mies, commander in chief of US Strategic Command, noted at a recent Congressional hearing that reduction of warhead numbers below 2,000 can lead to "unintended consequences."

"Tyranny of Numbers"

In his interview with NBC, Cohen elaborated on Mies's statement. "As you get to much lower numbers, you're looking at a tyranny of numbers," said the Secretary of Defense. "Namely, you could find yourself in a situation where you are forced to use it or lose it. ... It may force you to change your strategy as far as targeting, not strategic assets, but humans, which we don't want to do."

Often, this concern forms the basis of opposition to so-called "deep-cut" proposals, always plentiful in Washington.

A prominent supporter of deep cuts is Bruce G. Blair of the Center for Defense Information, a former Air Force officer and a longtime promoter of arms control agreements and reductions.

"The United States could easily drop to 1,500 warheads," he contended.

Blair observed that the US could hit that magic number by deploying 10 Ohio-class subs with only 480 Trident II missile warheads; only 300 single-warhead Minuteman IIIs; only 50 B-52Hs with 400 warheads; and 21 B-2s with 336 warheads.

"I think the Pentagon has overblown the difficulty here," Blair concluded.

Not surprisingly, DoD officials vigorously dispute Blair's methodology and conclusions. They note that Blair's force is able to achieve the 1,500-warhead number mainly by downloading each sub-based Trident II missile from five to two warheads. Problem is, START forbids downloading the missile below four warheads.

"It's not like we can just go to three [or fewer]," said a DoD official. "It's just not allowed."

Why? A nation that downloads a missile might be capable of reversing course and rapidly "uploading" a weapon to carry more warheads. "There's always uncertainty for the other guy," he added, "because there's space available."