Jeffrey LewisTerror Farm, Part 1

Peter Zimmerman and I have the cover article in Foreign Policy magazine (subscription only), arguing that nuclear weapons are a bargain for the large, sophisticated terrorist organization.

We basically asked, were we a terrorist group, how would we do it and how much would that cost?

Pete had previously collaborated with Nicolas Freeling to produce a novel about a would-be terrorist group building an HEU bomb, called Gadget. So, we spun out a similar scenario (with some important differences) at a ranch somewhere in Big Sky country, and then made some cost estimates.

I am pleased with the article, although anything written for a popular audience risks sounding a wee bit alarmist. The illustration of the ranch is titled “Terror Farm” which seems a little over the top, to me. (No, we didn’t write that.)

Overall, we tried to strike the tone of “This isn’t easy, but is also isn’t hard enough.”

The question of evaluating low-risk, high-impact scenarios like nuclear terrorism is always a difficult one. Bill Arkin has an essay in the Bulletin of Atomic Scientists that is probably a necessary corrective to some of the more alarmist expressions of the threat. Still, I’d sleep better knowing that global stocks of fissile material were under tighter control.

Pete and I will post a longer, scholarly version of the article (which is quite different and contains footnotes) at some point down the road. In the interim, I will probably blog about some of the ideas in the article that did not make the cut.

This is the first in a many part series based on my article with Peter Zimmerman in Foreign Policy entitled “Bomb in the backyard.” For previous posts click on the phrase terror farm.


  1. John Field (History)

    Jeffrey, I enjoyed the article and it’s obviously important for people in America to be thinking along these lines, but I think that a sophisticated terror organization faces additional cost constraints related to portfolio optimization.

    Essentially, this is a kind of speculative investing subject to cost(money/opportunities), return(casualties) and risk(getting busted).

    The problem is that it is very difficult to create a diversified “portfolio of terror” using nuclear weapons unless the organization is absolutely massive. It is analogous to an individual investing heavily in, say, iridium metal futures. It is irresponsible for an organization here for the long term to do this kind of one-off investing. I mean, it’s not like the war on terror is over after the bomb goes off, right?

    Ito’s lemma shows that the return on a random walk investment is reduced by half the volatility squared. Investing in the terror framework could be made mathematically precise starting in a stochastic calculus framework with Bellman’s equation and applying it to the decision path faced by a hypothetical terror planner. I believe such an analysis would show that smaller attacks like 9/11, LNG depots, chemical plants, etc. represent good investment choices for al Qaeda.

  2. abcd


    Congratulations on getting an article published in a magazine as influential and as large as FP. All of us wonkers here are very proud of our Dear Leader.

  3. xyz

    Decent article, but I’m not sure constructing a gun-type device is a particularly hard test for the theory that terrorists could nuke us. Fissile material still seems to be the major stumbling block, and you rather arbitrarily settled on $4 million (though I’m interested in seeing the longer version.) Given that al Qaida paid $1.5 million for a scam and that scams are part of the cost of doing business, just how many scams do you think al Qaida would have to sort through to get real material?

  4. Jeffrey Lewis

    Thanks for your comment — we picked the gun-type precisely because it was easy. I suspect a terrorist group would, too.

    I also agree with your obervation that scams are a part of doing business. In an earlier draft, we wrote:

    [G]iven our initial budget of $10 million, a well-financed terrorist organization need not worry unduly about scams or buying bogus NEM. The failed Al Qaida effort to purchase HEU from Sudan cost just $1.5 million – falling victim to such a scam would not threaten the cost effectiveness of developing a nuclear weapon. Poor quality or fake NEM may simply be an acceptable cost of doing business– provided the details of attempt to purchase nuclear material does not reveal the existence of the plot.

    And finally, the $4 million is way more arbitrary than I can tell you.

    Seriously, good comments.

  5. Pete Zimmerman (History)

    There are a lot of ways to cost out the HEU. You could use the cost of production of a critical mass in a standard commercial enrichment plant, and you would come in far below $4M. You could estimte the cost of bribes, hiring an insider, and actually stealing the stuff—or persuading the next AQ Khan to give you a deal. That might even cost less—or much more.

    What one cannot do is simply calculate based on what might once have been paid by Al Qaida for a scam.

    The existing “market price” is determined by a situation where there are crooks on one side and cops on the other. That’s not Adam Smith at work.

  6. Mike

    Yes, why worry about the real, existing nuclear weapons and the ‘statesmen’ who have been investigating making them more ‘userfriendly’ when we can scare ourselves with ridiculously theoretical scenarios about terrorists building nukes in basements?

  7. Alex (History)

    “Essentially, this is a kind of speculative investing subject to cost(money/opportunities), return(casualties) and risk(getting busted).” — and this is only really factoring in the risks/return of buying it, not the risk/return issue of selling it. How does one seek out terrorists to buy enriched uranium without seeking out a CIA operative at the same time? It’s the combination of these two factors which makes the idea of a fissile nuclear blackmarket sale seem improbable to me (though I haven’t been able to read the article yet, so I don’t know if you’ve discussed this or not in it—I’ll try to track down a hard-copy next week), or at least more improbable than, say, trying to seize the material from one or more research reactors which use HEU (the resulting bomb wouldn’t be efficient at all, but it could probably get the job done if destruction, death, and fear were your only goals). It seems to me that it would be pretty hard for a potential seller to find a potential buyer without an intermediary figuring it out and blowing the whole deal; at least, hard enough for me to think that other options would be pursued as far easier with probably similar political/cultural effect (i.e. IED’s in shopping malls and other cheap ways to scare and kill people).

    (But then again I’d be the first to admit that I’m not a policy person so much as a history person, much less a bomb engineer, so I don’t know if all of that adds up or not. And I might being displaying a phenomenal ignorance as to how easy it would be to get in contact with buyers if you were trying to actually sell fissile material).