At such point as the U.S. Senate begins to debate ratification of the new Russo-American strategic arms treaty ("New START"), most observers have assumed that fights will center around details such as its linkages to non-nuclear issues or the weakness of the verification regime it will install in place of the START agreement's elaborate processes. In light of the fact that the Obama Administration promised repeatedly that there would be no linkage in the treaty to missile defense, and that the administration's new Nuclear Posture Review (NPR) stresses the importance to U.S. strategy (and to the administration's disarmament agenda) of non-nuclear "prompt global strike" capabilities, the restrictions that New START places upon U.S. capabilities in these regards, and their significance, will surely be much debated.
Most commentators so far, however including me have assumed that the "top line" numbers in the new treaty will not prove terribly controversial in the Senate, if only because the reductions they entail are so modest (or perhaps even illusory). On its face, the fact that this much-hyped new agreement seemed to represent more of a whimper than a bang seemed to me likely to be a cause of major complaint only on the arms control left and therefore not a problem for Republicans and security-minded Democrats or Independents in the Senate. The ratification fight with them, I thought, would be held on other conceptual and textual terrain.
But that might not actually be the case after all. "New START" is apparently one of those agreements that looks stranger the more one stares at it ... and people are starting to stare. NPF will no doubt deal with the other issues (e.g., missile defense) some other time, but let's take a hard look at some of the "top line" numbers. I'm starting to intuit that they may not be so "easy" after all.
I. THE ARTICLE II(1) LIMITS
What precisely are the things that are limited? We've previously discussed the weird counting rules for deployed warheads, which are billed as a "one third" reduction but actually allow for the possibility that no cuts at all will occur under this treaty. But let's turn now to deployed delivery systems where it turns out there's a somewhat strange story too.
Let us, further, limit our discussion here to intercontinental ballistic missiles (ICBMs). Under New START's Article II(1)(a), the parties are limited to 700 "deployed ICBMs" each. What does this limit on deployed systems cover? The treaty Protocol helps explain this.
"Paragraph 37 of the Protocol says that "[t]he term 'intercontinental ballistic missile' or 'ICBM' means a land-based ballistic missile with a range in excess of 5500 kilometers."
"Under Paragraph 13 of the Protocol, "[t]he term 'deployed ICBM' means an ICBM that is contained in or on a deployed launcher of ICBMs."
"Under Paragraph 28 of the Protocol, "[t]he term 'ICBM launcher' means a device intended or used to contain, prepare for launch, and launch an ICBM."
Altogether then, if you have a land-based missile of over 5,500 kilometers in range in a device that can launch it, you have a unit that might fall within the 700-item cap on deployed systems. But not necessarily. Pursuant to Paragraph 28, if the launcher for this missile is not "deployed," this missile/launcher unit falls outside the limit of 700.
So far so good, I suppose. Everyone understands that non-deployed missiles are caught by the Article II(1)(c) overall cap of 800 for delivery systems, right? Well, not necessarily. This limit covers only "deployed and non-deployed ICBM launchers." Per Paragraph 48 of the Protocol, "[t]he term 'non-deployed ICBM' means an ICBM not contained in a deployed launcher of ICBMs." Remember, however, that Article II(1)(c) doesn't specifically cover "non-deployed ICBMs." It only limits "non-deployed ICBM launchers." Each party to the treaty can therefore apparently possess as many actual ICBMs as it wishes, as long as these missiles aren't actually in ICBM launchers.
(Missiles not in launchers are neither "deployed" nor counted as "non-deployed ICBM launchers" for purposes of the 800-unit cap for the simple reason that a "launcher" is not the same thing as a "missile," and no one apparently thought specifically to address non-deployed missiles in themselves at all. Under Paragraph 49 of the Protocol, "[t]he term 'non-deployed launcher of ICBMs' means an ICBM test launcher, an ICBM training launcher, an ICBM launcher located at a space launch facility, or an ICBM launcher, other than a soft-site launcher, that does not contain a deployed ICBM." Since Paragraph 13 defines "deployed ICBM" as one that is in a deployed launcher, these two provisions, read together, are a bit circular: a non-deployed ICBM launcher is one that does not contain an ICBM that is in a deployed launcher. And the term "deployed" is itself apparently nowhere defined. Welcome to treaty law.)
II. THE PROBLEM OF RELOADS
But what difference might it make that New START would seem to allow the unlimited possession of ICBM reloads? Well, for one, since such missiles are defined as non-deployed, there is no limit upon the number of nuclear warheads in them: the 1,550-warhead limit of Article II(1)(b) only applies to "warheads on deployed ICBMs" and non-deployed warheads are not constrained at all. Apparently, each side can have as many nuclear-armed ICBM reloads lying around as it wants and with as many multiple, independently-targeted re-entry vehicles (MIRVs) as its engineers can fit aboard them.
You might not think reloading would be much of an issue with respect to missile silos, and you might be right though one should remember that during the Cold War, Soviet strategists planned for protracted warfighting in part by preparing to decontaminate and reload such silos over the course of a few days. But what about mobile missiles of which we have none but which Russia is building as fast as it can and which are indeed capable of reload/refire operations? In previous arms control agreements, "rapid reload" was a recognized possibility, and it was understood that a mobile launcher might be reloaded in less than four hours. The 10,500-kilometer-range road-mobile SS-25 missile, for instance, is apparently capable of reload/refire operations.
If I'm reading "New START" correctly, then, an SS-25-style treaty-accountable deployed ICBM launcher could be accompanied by one or more nuclear-armed reload missiles and any necessary reload vehicles without such reloads being in any way limited by the agreement. That's good news for Russians, who are investing heavily in mobile land-based systems, but not so good if one is stuck with American-style silos.
III. NON-SELF-PROPELLED MOBILE SYSTEMS
But there's more. At least in this scenario a Russian road-mobile missile followed around by a reload vehicle with as many "war shot" reloads as one likes the mobile launcher itself is a treaty-limited item. (Article II(1)(c) constrains the total number of "ICBM launchers," whether deployed or non-deployed.) But a rail-mobile launcher that does not actually have a missile on it would not count as a "non-deployed mobile launcher of ICBMs" within the definition of this term provided in Paragraph 51 of the Protocol that is, "a mobile launcher of ICBMs that does not contain an ICBM" because pursuant to Paragraph 45, "[t]he term 'mobile launcher of ICBMs' means an erector-launcher mechanism for launching ICBMs and the self-propelled device on which it is mounted." (This assumed connection between self-propulsion and being a "mobile launcher of ICBMs" is repeated in the Protocol's Paragraph 57(c), which defines a "production facility" for mobile ICBM launchers as one at which "the erector-launcher mechanism of a mobile launcher of ICBMs is mounted on the self-propelled device.") A transporter-erector-launcher unit on a railroad car which is not self-propelled, for it must instead be pushed or pulled around by a locomotive would thus seem not to be caught by Article II limits on "non-deployed" systems at all.
Fretting about rail-mobile systems is no mere fantasy scenario, either. The enormous Soviet SS-24 missile, a MIRVed giant with 10 warheads similar to the U.S. Peacekeeper, was designed to be either silo-launched or rail mobile. The SS-24 was subject to elimination under the START agreement, but that treaty, of course, is no more and if you can build road-mobile systems like the SS-27, rail-mobile applications sound quite simple indeed. Rail-mobile systems are actually in no way unheard of elsewhere, either. India's Agni-1 shorter-range ballistic missile has been fired from a railway car; the Agni-III intermediate range missile has also been fired from rail-mobile launchers that press accounts describe as being "capable of launching the missile from anywhere in India." China has also studied rail-mobile systems in the past, and is rumored to be working on a rail-mobile platform for its new DF-31A as well. (NPF's counterparts at the Arms Control Wonk site suggest that such a Chinese railway TEL already exists.)
The impact of reload or upload operations upon strategic stability might not be so dramatic in the case of missile silos, because their locations are likely known, they are surely all closely watched in peacetime, and they would presumably be fairly vulnerable in a nuclear exchange. Stealthy uploading of empty silos would surely be very difficult, and wartime reloading highly problematic indeed. Mobile missiles, however, present quite another situation. If you can't find them, you cannot prevent uploading, or even count on knowing when the immediate strategic balance is being shifted against you through such activity. Furthermore, reloading in wartime remains a real possibility for mobile systems, and wouldn't take very long.
"New START" would seem to allow a party to have unlimited numbers of rail-mobile launchers deployed with nuclear-armed missiles, at least if these missiles are not actually uploaded. In such a scenario, if you gave the signal, all your rail-mobile launchers could be uploaded with MIRVed missiles in extremely short order, but prior to that point, none of these launchers or missiles or warheads would fall within treaty limits: you could have as many as you like. With a suitable investment in such systems, in other words, one could lawfully linger only a few hours away from a sort of "add-water-and-stir" strategic breakout capability. Since Russia is presently building mobile ICBM systems and we are not, some Senators might find this possibility worrisome.
And the situation could actually be worse than that. Article II(1)(c) of the "New START" agreement, as we have seen, limits "deployed and non-deployed ICBM launchers." Since an "ICBM launcher" is defined by Paragraph 28 of the Protocol as "a device intended or used to contain, prepare for launch, and launch and ICBM," you might think that a rail-mobile TEL with an uploaded missile would count under this limit.
Weirdly, however, it may not be that simple. As we have seen, Paragraph 45 of the Protocol defines "mobile launcher of ICBMs" in terms that would not reach rail-mobile (or any other non-self-propelled) launchers. The Protocol's definitions, in Paragraphs 16 and 51, for "deployed" and "non-deployed" mobile ICBM launchers piggyback upon Paragraph 45. (This is hardly surprising: if it's not considered a mobile ICBM launcher in the first place, it cannot be either a deployed or a non-deployed mobile ICBM launcher.)
Here's the trick: it's not entirely clear that rail-mobile systems or any other non-self-propelled mobile systems, for that matter fall within the Article II(1)(c) cap at all, even if uploaded with missiles. As noted, based only upon the language of Article II itself, one could make a decent argument that they do. But the Protocol complicates things. It does not define "deployed ICBM launcher" or "non-deployed ICBM launcher" in themselves, thereby leaving inexplicably undefined the basic categories that are constrained by the Article II(1)(c) cap of 800 units. It does, however, specifically define what counts as a deployed or non-deployed mobile launcher. This suggests that the treaty should be viewed as not subjecting to Article II(1)(c) limits anything that is mobile but doesn't fall within the "mobile launcher of ICBMs" definition of Paragraph 45.
And that may not be all. The Protocol's Paragraphs 13 and 14 define "deployed ICBM" as an ICBM that is "contained in or on a deployed launcher of ICBMs" and define "deployed launcher of ICBMs" as an ICBM launcher that contains an ICBM and isn't a test or training unit, and isn't at a space launch facility. If one is to infer from the Protocol's strange construction that mobile non-self-propelled systems don't count as ICBM launchers in the first place for, as we have seen, they are not reached by the Protocol's definition in Paragraphs 16 and 45 of deployed mobile launchers, and they certainly aren't immobile such devices may not even fall within Article II(1)(a)'s 700-unit cap on "deployed ICBMs" either.
If this is the case, rail-mobile or other non-self-propelled mobile ICBM systems wouldn't be limited by the treaty in any way, whether or not they are deployed or uploaded with missiles. This would presumably be good news for the growing Russian mobile missile industry, but hardly for the Americans that would be targeted by such devices which Russia would seem free to possess in any numbers it likes.
Perhaps an attentive NPF reader will be able to convince me that I've misread "New START" and its protocol. I hope so: the holes the agreement seems to leave in these definitions appear greatly to favor Russian strategic planners over their American counterparts, and it would bode ill for Senate ratification if the members of the Obama Administration's varsity negotiating squad had been so out-maneuvered by their Russian partners. (For the record, NPF would be delighted to post on this website an administration rebuttal or clarification on these points, and I invite such feedback. I may be reached at ford@hudson.org.)
One could perhaps argue about whether the United States really needs a traditional Cold War-style arms control treaty with Russia at all at this point, though it has been a centerpiece of the Obama Administration's policy that we do. If we are to have a treaty, however, surely it should be one that lacks major loopholes.
Unless I'm greatly misreading the text, therefore, don't count on the treaty's "top line" being found quite so easy after all.