Jeffrey LewisDead Hand, START and Strategy Stability

Last week, the Senate Republican Policy Committee released this execrable document, START: Do Time Extension Instead of a Bad Treaty, on the New START treaty.

Today, I finagled a copy of a memo that Senator Kyl’s staff has distrubuted in advance of a briefing that Assistant Secretary Rose Gottemoeller will give to the Senate National Security Working Group tomorrow at 9 am.

The memo, October 8, 2009 briefing with START negotiating team, is not very encouraging.

Taken together, these two documents are disheartening, in that they depict an opposition to the President that isn’t serious.

So, this is sort of a long post on why the New Strategic Arms Reduction Treaty (New START) is important to US national security, but it is also in some sense an elegy to the dwindling number of moderate Republicans who played such an important role in setting US nuclear weapons and arms control policy.

Arms Control Is A Favor We Do Ourselves, Not The Russians

I want to start by addressing the juvenile tone of the Senate Republican Policy Committee document, which repeatedly implies the new START treaty is some kind of favor to Moscow.

The document repeatedly asks whether Russia has “earned” further reductions or whether its behavior “warrants” a new treaty.

This is crazy. The Administration seeks a New START agreement for the same reason that a McCain Administration would have: because it is in our interest, for at least two reasons.

1. A new START is important to drive the Russians toward a more stabilizing strategic posture that does not depend heavily on MIRV’ed ICBMs, and

2. A new START is essential to our ability to monitor Russian nuclear weapons programs.

I want to talk about each of these in turn, but indulge me for a bit on the subject of strategic stability.

Strategic Stability

A lot of people were shocked to hear Secretary Gates explain (in an unclassified setting) that one of Moscow’s main concerns was that “ground-based interceptors in Poland could be fitted with nuclear weapons and become an offensive weapon like a Pershing and a weapon for which they would have virtually no warning time.”

Yes, I know that sounds very War Games. But the Russians have always been paranoid about decapitation strikes against their creaky command and control structure, from the War Scare of 1983, through the 1995 Black Brant fiasco, right up to today.

Understand, this is the strategic culture that gave the world “Perimeter” — the so-called “Doomsday Machine” or “Dead Hand” detailed in David Hoffman’s book of the same name and in a recent Wired article by Nick Thompson. Perimeter, as Thompson writes, was largely a measure to compensate for inadequate Soviet command-and-control capabilities.

The fact that Moscow still worries today about command performance should give us pause. It is not in our interest for the Russians — and their giant nuclear arsenal — to operate on the basis of paranoid fantasies about the United States.

Arms control is one way to address that. So let’s be clear, we do this because it is in our interest.

Russia’s Declining Strategic Forces

The Senate Republican Policy Committee asks what is intended to be a rhetorical question: “Why pay for what is free?” Russia’s strategic forces are in decline, so why agree to cut ours?

Over the next decade, in the absence of any arms control treaty or agreement, “the number of delivery vehicles in Russia’s nuclear arsenal will continue to decline sharply,” perhaps to fewer than 500 delivery vehicles.55 This is because “Russian strategic systems have not been designed for long service lives,” and Russia is unable to replace aging delivery systems at the pace at which they are retired.56 There is certainly no reason for the United States to pay for something that is going to happen with or without an arms control treaty. In this respect, there is no reason for the United States to sacrifice U.S. nuclear force structure, or other
unrelated national defense matters, such as missile defense or prompt global strike, “as a price to be paid for an agreement that requires nothing of the Russians beyond discarding the aged systems they plan to eliminate in any event.”57

This is fantastically misleading!

What is declining in Russia is the number of delivery vehicles — missiles, bombers and the like — not the number of warheads.

The Republican Senate Policy Committee attempts to obscure this distinction by stating that “Russia needs this agreement far more than the U.S. does. It is desperately trying to lock the U.S. into lower nuclear levels, not the other way around.”

“Nuclear levels” is a meaningless phrase intended to deceive, not elucidate — Russia wants lower numbers of delivery vehicles but more warheads. The United States seeks the opposite: lower levels of warheads — much, much lower than even in the Joint Understanding — but many more delivery vehicles.

Senator Kyl’s letter acknowledges this, noting that “the Russians have been testing a new multiple-warhead version of the Topol-M ballistic missile” that would be prohibited under START.

Russia is deploying the MIRV’ed Topol because Moscow wants to keep its warhead numbers constant, even as the number of delivery vehicles plummets.

In other words, Russia is heading toward MIRVing the hell out of its strategic forces to keep its warhead numbers up around 1700. This is probably the only thing I agree with in Senator Kyl’s letter — I am also disturbed by the deployment of the MIRV’ed Topol. I don’t worry that the missile itself disturbs the strategic balance, but I do worry about what the MIRV’ed Topol deployment says about trends in future Russian strategic forces.

Recall the discussion of nuclear decapitation and strategic stability. It is not in our interest for Russian leaders to be confused about the possibility of a decapitating U.S. first strike — unless the thought of a half-drunk Boris Yeltsin staring at the Russian “football” doesn’t bother you. Now, imagine that a significant fraction of Russia’s nuclear forces are deployed on a small number of relatively vulnerable land-based ballistic missiles. What impact do you think that will have on the time pressure faced by a future Russian leader?

A very senior Bush Administration official, one who was deeply involved with negotiating the original START, once said to me: “One way to look at the arms control endeavor is as a bipartisan effort over the past thirty-years or so to drive the Soviets and now the Russians to a more stable strategic posture.”

That really stuck with me, because I think it is dead-on. Indeed, Kerry Kartchner’s history of the START negotiations, Negotiating START: Strategic Arms Reduction Talks and the Quest for Strategic Stability, makes this point eloquently:

The U.S. approach to START was, above all, a quest for strategic stability, the Holy Grail of the nuclear era. In fact, early Reagan statements made it clear that no agreement would be better than an arms control accord that failed to enhance strategic stability. This pont underscored the view that arms control was a means to an end, and certainly not the only means …

This is why, for instance, the George H. W. Bush Administration negotiated a START II agreement with Russia that banned MIRV’ed ICBMs.

Obviously, I’d like to get back to that MIRV ban. (MIRVs and the Moscow Treaty, December 12, 2004). But that’s not going to happen (STRATCOM Hearts MIRV, January 30, 2006, STRATCOM Still Hearts MIRVs, November 29, 2007).

In the interim the best we can do is try to make sure the New START agreement strikes a better balance between the number of operationally deployed warheads and the number of strategic delivery vehicles than what we are likely to have absent an agreement, while preserving the essential monitoring and verification provisions.

Why Not Just Extend START for 5 Years?

Let me begin by saying that I favored START extension throughout the twilight of Bush Administration.

I tried puns (START Talking March 7, 2007), the obvious (Extend START, April 20, 2007), and the over-the-top (Frickin’ Extend START Already, June 21, 2007) before slipping into despair (START: Dead Treaty Walking, September 22, 2008).

But the Bush Administration, or at least the parts that mattered, wasn’t interested because of the paperwork burden imposed by the verification measures. (Sadly, that is not hyperbole.)

So, believe you me, when Senate Republicans suddenly say “We should just extend the treaty,” well that comes at it mighty high.

The reality is that “extending START” is, as options go, a poison pill. If Rose Gottemoeller shows up in Geneva and says “scrap the joint understanding, let’s just extend START,” the Russians are going to take their MIRV’ed Topol ICBMs and go home.

We had a chance to extend the START Treaty — which I favored — but that opportunity is now past.

What about Verification?

Finally, Senator Kyl’s memo states that a draft New START “was not accompanied by the important verification protocol.” The implication is that the Administration has no intention of negotiating verification measures, which I sincerely doubt.

Again, given that the Bush Administration negotiated the Moscow Treaty without a verification protocol, this is somewhat churlish.

But the fact is that maintaining the verification and monitoring provisions in START is an important interest. We know, thanks to Jonathan Landay’s excellent reporting, that the Intelligence Community issued NIE on Russian strategic forces that expressed doubt about our ability to monitor Russian compliance with the Moscow Treaty without the measures in the START agreement. (See: IC Can’t Verify Moscow Treaty, December 22, 2004).

So, I guess this is the second area in which Senator Kyl and I agree — the verification provisions will be important. I am a little surprised to see this, because the Senate Republican Policy Committee was rather blase about the demise of START noting “If the verification regime is extended, both Russia and the United States benefit similarly; whereas if it lapses, there is probably equal detriment.”

That’s asinine.

Let me ask you: Do you think it is easier to make reliable open source estimates of US or Russian strategic forces? To say that the lapse of the verification provision would be to the equal detriment of both sides is one of the more foolish things I’ve seen in a long time.

Please Be Serious

This is really a plea for Senate Republicans to play a constructive, engaged role, rather than being arm-chair negotiators. (Hey, that’s my job!)

I happen to think that the START process under Reagan and Bush greatly improved on SALT, not the least for the conceptual approach taken by self-consciously “conservative” supporters of arms control that emphasized strategic stability over conviviality. In retrospect, I think the START I and START II treaties were impressive pieces of work that reflected both Democratic and Republican priorities.

I suppose it only seems that way with hindsight. But, right now, I am feeling nostalgia for the good old days.


  1. Pavel (History)

    Jeffrey: Topol-M is not a MIRVed version of Topol – these are different missiles. RS-24 is a MIRVed version of Topol-M. (If those Russian designations sound confusing it is because they are.)

  2. Scott Monje (History)

    “It is not in our interest for the Russians — and their giant nuclear arsenal — to operate on the basis of paranoid fantasies about the United States.”
    I agree with this, and yet it strikes me as interesting that so many people (certainly not just you) describe the Russians as “paranoid” when they are basically pursuing the behavior predicted by standard realist theories. We, on the other hand, decided as soon as the Cold War was over that the only country in the world capable of obliterating the United States was no longer worth thinking about and then proceeded to obsess about secondary powers that just might one day maybe get one bomb. (The Russians shrug and say that if Iran ever does get an ICBM the Americans will just bomb the launch pad.) It’s really one of the most devastating blows to realist theory that you can imagine.
    I would just add a couple of points to your comments. First, the Russians want the New START to permit many warheads but few delivery systems to force us to pile up the warheads and eliminate any sudden upload potential. Second, MIRVed missiles are destabilizing not only because they reduce the number of targets, but also because one missile can take out several. Thus even if both sides have equal numbers of missiles and kill probabilities are less than 100%, the side that shoots first can still take out the other side.

  3. anon (History)

    I’m not sure where to begin (or end) with comments on both the Republican’s report and your comments. I may come back to this a few times during the day, as I chew on this and think of more stuff to say.

    Two comments to start with:

    It is well worth remembering (and it seems that Kyl has forgotten) that START is a 5-party treaty. The United States can not simply decide to extend the Treaty for 5 years, it needs the approval and agreement of the other 4 parties. Russia does not want to extend the Treaty, in part, because it does not want the other 3 FSU parties to have any role in the future of arms control. If the U.S. changes course now, and asks for an extension, then it will be admitting that the treaty will lapse and be replaced by nothing. Also, the absence of support for extension is not simply a reflection on the paperwork associated with the verification regime. Russia wants to deploy a MIRVed Topol; it can’t do that under START and therefore does not want to extend START.(Kyl’s letter gets at this. Yes, Russia is gaming the clock on the Topol, but its gaming for START expiration, not for a new START Treaty. Its been playing this game for over a year now, even when the Bush folks had no interest in a new START.) The U.S. also feels pinches from the Treaty, and wants out of some of the provisions. So, if you ask the U.S. military, you’ll hear that a 5 year extension may do more harm to U.S. security than a new Treaty that releases the U.S. from some of its START obligations.

    Second point, on strategic stability. I’ll agree fully with the Republican report that the U.S. should only sign a new Treaty if it serves U.S. security interests. But I’ll disagree fully with their attempt to define “serving U.S. security interests” as the equivalent of “reducing Russian delivery vehicles.” Yes, we will get a reduction of delivery vehicles for free. As you said, this may do more harm than good for strategic stability. But its also not the right measure of whether the treaty is in U.S. security interests. The Treaty puts U.S. and Russian nuclear weapons in a fixed box. The size of the box does not matter as much as the existence of the box. The box provides predictability and transparency for both sides, reducing the likelihood of misunderstandings and worst-case assessments. It also calls for extensive cooperation in monitoring and verification, which can build understanding and enhance confidence. The Treaty steps both sides back from a nuclear precipice. The Bush Admin did not think this was necessary because it thought we were far enough off the precipice anyway; if we still see risks from our nuclear relationship with Russia, then this Treaty can help reduce those risks. I really, really want to hear the Treaty’s supporters start to argue the case for the treaty on enhancing U.S. security, rather than on re-setting U.S.-Russian relations.

    A few more points (I know, I said I’d start with 2, but I’ve thought of more.) Kyl complains about Russian compliance with START and about the lack of a compliance report. Its worth remembering that we had only one (rather than one per year) during the whole Bush Admin because Paula DeSutter didn’t want to produce them (or tried to produce reports that were so far out in right field that she couldn’t get agreement). He’s complaining about a problem created by the a Republican Administration. Also, on verification, its likely that the Republicans will shoot massive holes in whatever verification regime new START has, and will be particularly scathing about the absence of provisions that were included in the original START. But this is the same team that accepted, without question, the entire absence of verification in the Moscow Treaty and is quite willing to let START, and all its verification expire. Hard to argue that verification is the key to new START when you were willing to live without any verification…

    Final point; the report harps on Russia’s nonstrategic nuclear weapons and insists that a new Treaty address the disparity. If you go back to the hearings on the Moscow Treaty (and even on START II) you’ll find that only the Democrats were concerned about Russia’s nonstrategic nuclear weapons. Rumsfeld and Powell admitted we didn’t address them in the Moscow Treaty, and said maybe we could talk about them in the working groups that we had agreed to have with the Russians. The Republicans didn’t seem to care. Now they care, but now the discrepency in U.S. and Russian nonstrategic nuclear weapons is actually smaller than it was in 2002. Go figure.

    I’ll stop here. Thanks for letting me get this off my chest….

  4. Chris Ritter (History)

    An “execrable document” indeed, Jeffrey, which begs yet more criticism. Which I will forgo. But I have to ask about the last of its “principles”:

    6. A comprehensive nuclear modernization plan should accompany the treaty.

    Is this an attempt to revive something from the good (not-so-)old days for which none should have the slightest nostalgia, the RRW?

  5. Stephen Young (History)

    Another sad fact – the Republicans aren’t even listening to the documented advice of their own wise men, those Cold War-era nuclear true believers who represented half of the Congressional Commission on the U.S. Strategic Posture, which gave a clear-headed, strong endorsement of a new treaty in their final document: “The moment appears ripe for a renewal of arms control with Russia, and this bodes well for a continued reduction in the nuclear arsenal. The United States and Russia should pursue a step-by-step approach and take a modest first step to ensure that there is a successor to START I when it expires at the end of 2009.”

    (Ignoring, for the moment, some of the less supportive comments a few members have made since that public statement.)

  6. Jeffrey Lewis (History)

    Pavel, thanks — I was more careless than I should have been. Have updated the Topol references to make clear the issue is that it is a MIRV’ed Topol that is the object of concern.

  7. anon

    I am sorry, but I see no legal reason why you still couldn’t extend the treaty for 5 years by its plain terms. Anybody here disagree? I understand the policy reasons for more than mere extension, but I failed to see the “it’s [now] impossible” reasons—legally speaking.

  8. Kingston (History)

    Re: Stephen’s point about the RPC’s selective reading of the Strategic Posture Commission report, it’s important to note that while the RPC cites the Commission (by my count) 25 times, it omits the one Commission statement that is most relevant, which is the one Stephen cites above. For more on this, see here:

  9. Scott Monje (History)

    “Russia wants to deploy a MIRVed Topol; it can’t do that under START and therefore does not want to extend START.(Kyl’s letter gets at this. . . .”

    START did not ban MIRVs. On the contrary, START permitted 6,000 warheads (4,900 attributed to ballistic missiles) and only 1,600 delivery vehicles, virtually enshrining MIRVs. START II banned MIRVs (only on land-based ICBMs), but that never went into effect.

  10. Jeffrey Lewis (History)


    START did, however, prohibit “increasing the number of warheads attributed to an ICBM or SLBM of an existing or new type…”

    Topol-M is attributed 1 warhead under START.

    Hence, the Russian claim that the RS-24 is a “new” missile, even though it doesn’t qualify as a “new type” under the START agreement.

    Pavel explained this very clearly in this post.

  11. Pavel (History)

    Jeffrey: Did you mean “MIRV’ed Topol-M”? 🙂 Speaking seriously, it is quite possible that RS-24 can be declared a new missile – if it’s throw-weight is declared to be 1250 kg and not 1200 kg of Topol-M -it’s the 21% increase in throw-weight rule (it’s not 1210 kg because TW is declared in increments of 50 kg). Given the rounding, the actual trow-weight could be, say, 1226 kg. I’m sure they could scrape up a few kilograms of throw-weight if they want to. I was told it’s not a big deal.

  12. Rhyolite (History)

    “ground-based interceptors in Poland could be fitted with nuclear weapons and become an offensive weapon like a Pershing and a weapon for which they would have virtually no warning time.”

    Couldn’t the same kind of decapitation without warning be achieved with a low observable aircraft like the B-2 or a low observable cruise missile like the AGM-129? A ballistic missile attack of any kind would seem inferior in this regard.

  13. Anon.


    First, your math does not compute. 21% of 1200kg is 252kg, not 25.2kg. The russians would have to reach a throw weight of 1450kg to 1500kg (depending on how it is rounded to a multiple 50kg) to reach that limit. Doesn’t sound to me like scraping up would do.

    Second, I have understood (although I may be in error here, I’m no expert in legalese, even less so in treatyese…) that the throw weight rule works the other way around: New missiles can have any throw weight, but improved versions of existing missiles may not exceed 20% extra over the original version.

    Third, I have understood that the throw weight considerations are moot as the RS-24 is concerned. Since the treaty forbids increasing warhead counts on existing designs, the “MIRVed Topol-M” must be declarable as a new missile, otherwise it is forbidden by the treaty, even if it’s throw weight was exactly 1200 ounces.

  14. Pavel (History)

    Anon: Topol/SS-25 is the original missile, with throw-weight of 1000 kg. So, 21% is 210 kg.

    On your second and third points – no, you got it wrong.