One of the more tiresome points being made in relation to the revelations from Edward Snowden is that there is nothing really new here. And, of course, it’s true that, if you’ve been paying careful attention to all the news on this topic, disregarding both official assurances and the wilder conspiracy theories, and thinking through the implications, the material leaked by Snowden is more confirmation than revelation. But, sad to say, that’s not the case for most of us. I think I’ve been paying more attention than most, and I still learned a lot from the latest news.
That’s all a preamble for a repost of a piece I wrote in 2004, in relation to an earlier revelation along similar lines, with a link to an even earlier piece from 2001, making the general case that secret intelligence is useless.
Gentlemen don’t read other gentlemen’s mail (repost from 2004)
The news that British spies bugged the office of Kofi Annan during the Iraq debate has a number of implications. First, for me, this is the point at which Tony Blair should go. The whole idea of going to the UN for authority to invade Iraq was his, not Bush’s, and now it’s clear that it was corrupt from the beginning. I won’t argue this in detail – no doubt a lot of people already thought he should go, and others still won’t be convinced.
The main point I want to make is that it’s time for Britain to get out of the spy game. More than any other democratic country, Britain is addicted to spies and their natural counterpart, Official Secrets. From Burgess and McLean to the present day, the spies have been a constant cause of embarrassment and worse. On the other hand, there’s no evidence that they’ve ever found out anything that was both useful and sufficiently reliable to act on (in this context, I’m excluding wartime codebreaking, which is always useful since, at a minimum, it disrupts enemy communications).
This isn’t a matter of bad luck, or even incompetence. Standard game-theoretic reasoning shows that, outside the zero-sum case of war, there’s unlikely to be a net benefit from actions like bugging offices. The problem is simple. If I bug your office and you don’t suspect me, I can gain potentially valuable information that you don’t want me to have. But if you suspect me, and I don’t suspect that you suspect, you can use my bugs to mislead me. As with all game theoretic reasoning, you can iterate this as many times as you like, but the end result is that the net value of information derived from bugging is zero. On the other hand, the costs of the activity are substantial. In an environment where bugging is routine, everyone learns to communicate in various forms of code, and decoding is costly and prone to error.
He’s often been dismissed as hopelessly naive, but US Secretary of State Henry Stimson was right when he shut down the State Department’s cryptanalytic office saying “Gentlemen don’t read each other’s mail.”
h4. Notes
I’ve treated Clare Short’s allegation as fact, since Blair hasn’t denied it. His claim that he can neither confirm nor deny it for security reasons doesn’t hold up. Short made the specific claim that she had seen transcripts of Annan’s conversations. Blair could refute this claim without reference to whether or not such transcripts existed.
The argument about the uselessness of spies is developed at much greater length in this piece I wrote for the Australian Financial Review The conclusion:
The spy myth clearly served the interests of intelligence agencies, which prospered during the 20th century more than any set of spies before them. The real beneficiaries, however, were the counterintelligence agencies or, to dispense with euphemisms, the secret police, of both Western and Communist countries. The powers granted to them for their struggle against armies of spies were used primarily against domestic dissidents. Terms such as ‘agent of influence’ were used to stigmatise anyone whose activities, however open and above-board, could be represented as helpful to the other side.
The supposed role of the secret police, to keep secrets from opposing governments, was, as we have seen, futile. Secret police, and the associated panoply of security laws, Official Secrets Acts and so forth, were much more successful in protecting their governments’ secrets from potentially embarrassing public scrutiny in their own countries.
As spies and the associated fears have faded in their public mind, their place has been taken by terrorists. In many ways, this is a reversion to the 19th century, when the bomb-throwing anarchist was a focus of popular fears and the subject of novels by such writers as Chesterton and Conrad.
As the attacks of September 11 showed us, the threat posed by terrorists is real. Nevertheless, even if terrorists were to mount attacks ten times as deadly in the future, they would still present the citizens of the Western World with less danger than we accept from our fellow-citizens every time we step into our cars.
If the century of the spy has taught us anything, it is that we need to assess the dangers posed by terrorists coolly and calmly rather than giving way to panic.
{ 104 comments }
leederick 06.23.13 at 10:20 am
I know the focus is on diplomatic espionage, and I think you are right in that context. But economic espionage – which hasn’t been discussed much , but is certainly happening – could be hugely profitable.
Barker 06.23.13 at 10:31 am
But how do I know you’re a gentlemen if I don’t get to read your mail?
sean matthews 06.23.13 at 12:19 pm
There was a very nice article a year or so ago about that famous counterintelligence operation with the washed up body on a Spanish beach with the briefcase full of documents that the British organised during the second world war. It makes precisely your point about spying. The usual form of the story is that the Germans snuck a look at the documents, and drew the intended (wrong) conclusions about allied intentions, and that this was a triumph of brilliant british counterintelligence. The reality is that everybody involved was second, third and fourth guessing everybody else, and the Germans quite explicitly considered the possibility that the whole thing was a planned hoax – they appear to have jumped to one set of conclusions rather than another only by a hairsbreadth of conviction.
Good article – worth tracking down.
Consumatopia 06.23.13 at 1:29 pm
The game theory of counter-terrorism might be different than that of obtaining intelligence about a hostile foreign government–if you find out someone is planning something, you don’t care precisely what they’re planning, you just go capture/kill them. You might get the wrong person, but at least that wrong person isn’t themselves the one trying to trick you.
Arguably the CIA is doing a better (i.e. less terrible) job of counter-terrorism today than they did of intelligence and counter-intelligence in the Cold War. Of course, it would be hard not to.
Andrew F. 06.23.13 at 1:58 pm
But does the reality of intelligence work reflect the assumptions structuring the theoretical game in the post? This reminds me of the discussion recently about an economics article on an asymmetric equilibrium between states like the US and states like Sweden: the assumptions are important, and they may not apply.
To give one possible example, party A may have both better ability to detect disinformation from party B and better ability to discover information from party B. The outcome and payoffs of the game may depend significantly on who is playing.
peter 06.23.13 at 2:06 pm
sean matthews # 3:
Regarding Operation Mincement, the 1943 counter-intelligence case you mention: The Abwehr, German military intelligence, was led, from the top down, by committed anti-Nazis (many of whom were later executed by the Nazi regime). The interpretation of the Mincemeat events finally adopted by the Nazi regime (the interpretation intended by the allies) may have been subtly promoted by the anti-Nazis in the Abwehr.
Likewise, Britain’s great success in turning every single Nazi agent sent to Britain during WW II who was not killed into a double-agent sending false message back to Berlin may have been greatly helped by the fact that the sponsors of these agents in Berlin and elsewhere in Europe were people in the Abwehr.
peter 06.23.13 at 2:35 pm
Game theory, which assumes people act in their own self-interest and play by some known, shared rules, strikes me as particularly useless here. I think there are two flaws in this game-theoretic argument against bugging:
The first: this is an argument that, after infinitely many iterations of reasoning about the other player’s reasoning, the value of information to the party doing the bugging will be zero. But, perhaps unlike economists, most of us live in a finite world. At every finite step along the way before we have iterated infinitely often, the value of information may be non-zero, even when this value approaches zero at infinity (and even when it approaches zero quickly). Convergence to a number is not the same thing as ever being equal to that number, a flaw in reasoning that took pure mathematicians a whole century to sort out in the differential calculus (roughly 1750 to 1850).
A similar argument is used by epistemic modal logicians to demonstrate that common knowledge is not common insofar as there is any finite depth of knowledge about the another party’s knowledge which any party does not know – see Michael Chwe’s book “Rational Ritual” for an easy introduction to these ideas.
The second flaw is that this game-theoretic argument is an argument about information, verifiable facts about the world. Much of espionage – perhaps most of it, and certainly all the most valuable parts – is not about collecting verifiable facts about the world, but in discovering other people’s intentions and predispositions. These mental states are rarely verifiable (they are sometimes not even known to the protagonists) and they are subject to all manner of paradoxes, quirks and manipulation, from self-fulfilling and self-denying prophecies to the paradoxes of public announcements. Two years ago, the motorists of Britain heard a Minister for Transport saying there was no shortage of petrol but he would be filling his jerry-can just in case. Within hours, petrol stations across the country had long queues, suburban London was enmeshed in hours-long traffic jams, and there was indeed a shortage of petrol.
Watson Ladd 06.23.13 at 4:10 pm
I can think of three recent cases where intelligence was useful. The first was Olympic Games: sabotage of the Iranian nuclear program possible only because of knowledge of how it was progressing. That sabotage may have prevented a war. The second is the Airbus Boeing bidding in which Boeing obtained information from US intelligence that Airbus was bribing officials, and used it to disqualify the bid. The third was finding Osama bin Laden: knowing he used a courier, and following the courier, revealed his location. Clearly, something is wrong with the argument presented here.
Tim Wilkinson 06.23.13 at 6:32 pm
peter: Game theory doesn’t really assume people ‘act in their own self-interest’ nor, more importantrly, that they ‘play by some known, shared rules’. Its application assumes that they have interests which can be modelled, and that they play by rules imposed by their situation and known to the modeller, but that is not the same thing. Certainly JQ’s specific point depends on some minimal assumptions about the spy agency’s capacity for reasoning and its beliefs about the opposition, but those assumptions 1. aren’t a part of the structure of game theory and 2. are realistic, indeed true.
Just as homo oeconomicus – and the weak EMH – is not that inaccurate a model when applied to n-c economists’ natural constituency, traders on the stock market (which is not to say that the alarming instability of an iterated Keynesian beauty contest has been adequately modelled), so the model of sneaky quasi-paranoid strategists second-guessing each other to arbitrary depth is quite an accurate model of spy vs spy. And that second-guessing doesn’t need to be iterated to any very great depth before the value of such espionage drops to a value close enough to 0 as makes no difference for cost-benefit purposes.
Thanks for the pointer to Chwe(any public internet source would be much appreciated). The point about common knowledge may (IIRC) follow from Lewis’s explication/definition, though we could certainly get into arguments about transmission of warrant v closure under known implication and analogues.
But in any case, I think the whole thing becomes intractable after rather few iterations. To say that I know that you know that I know that p, and vice versa, is just about possible to follow, but how much sense – and use – can we really make of the proposition that you don’t know that I know that you know that I know that you know that I know that you know that I know that you know that p? Well before we get to that point, no-one really knows what is going on, which I take to be part of what JQ is arguing. I certainly don’t see how this kind of thing is going to provide a refutation of his conclusion.
Your second, quasi-Heisenbergian, point seems to support JQ’s contention that the whole thing rapidly devolves to uncertainty. Incidentally, I don’t agree with the metaphysical claim that intentions (e.g. plans) don’t count as ‘facts about the world’, but that doesn’t really affect the issue: I take it that the operative idea here would be that intentions can be changed if it’s thought that they are known to others; but the same goes for undeniably real and objective things like passwords or the location of military units: in any case, this possibility only goes to support the contention that the whole spy vs. spy game has little value.
hix 06.23.13 at 7:28 pm
One has to bear in mind just how high the financial hurdle is that has to be overcome before any net gains would emerge, when spying is done on such a large scale. The US government intelligence budget amounts to at least 75 billion or 0,5% of gdp.
http://www.examiner.com/article/panetta-reveals-military-intelligence-program-budget-requests
http://www.dni.gov/index.php/newsroom/press-releases/96-press-releases-2012/756-dni-releases-fy-2012-appropriated-budget-figure?highlight=YTozOntpOjA7czo2OiJidWRnZXQiO2k6MTtzOjk6ImJ1ZGdldGluZyI7aToyO3M6NzoiYnVkZ2V0cyI7fQ==
Stephen 06.23.13 at 7:52 pm
JQ may not be entirely surprised when I say that I only partly agree with him.
That much intelligence activity has been futile: no argument there. Though much of human activity is ultimately futile, even teaching economics, intelligence activity may be more so than usual.
That Tony Blair was utterly wrong in pressing for the invasion of Iraq in the way he did: no argument there either.
That bugging Kofi Anann’s office made the idea of going to the UN for authorisation “utterly corruptâ€: sorry, can’t follow that argument. Part of a massive deception, yes.
That “wartime codebreaking … is always useful since, at a minimum, it disrupts enemy communicationsâ€: a wonderful understatement. At some very considerable way above the minimum, there is no doubt at all that wartime codebreaking revealed the Zimmerman telegram, allowed the US to identify Midway as the main Japanese target, helped enormously in the defeat of the WW2 U-boat offensive, and to cut Axis supply lines to North Africa … I could go on.
That “there’s no evidence that [spies have] ever found out anything that was both useful and sufficiently reliable to act onâ€: well, there is a problem here, in that secret services’ activities are by definition secret, so certainty about exactly what was found out is hard to achieve. Still, if reports are approximately accurate, it does seem that:
Richard Sorge found out that Japanese war plans in 1941 would be against the US and UK, not against Russia. That is supposed to have encouraged Stalin to move Siberian divisions westwards, hence the German defeat at Moscow. I would reckon that as useful; Soviet/Nazi regimes both awful, but victory for latter worse from point of view of liberal civilisation.
The Rote Kapelle organizations inside Germany and Switzerland provided very useful intelligence to Moscow. Also useful.
Klaus Fuchs and others conveyed information about atomic bombs to the Soviets. Some would argue this useful also: just think what the unspeakable proto-fascist US government might have done with a nuclear monopoly.
Penkovsky told the US about Soviet missiles in Cuba, in time for the US to get them removed. Likewise arguable.
Gordievsky explained to the UK that the Soviet government were interpreting exercises like Able Archer as preparations for attacking the USSR, and helped to prevent WW3 breaking out by accident. Useful, no?
Efforts by the UK, Irish and American intelligence services to bug and infiltrate Irish Republican organizations led to Adams, McGuiness and others being of great service to Her Majesty’s Government, in bringing the late Troubles to a close, with Northern Ireland remaining part of the United Kingdom in accordance with the democratic wish of most of its inhabitants. Useful from most points of view, also.
I won’t go into deception operations (Mincemeat, FUSAG, etc) here, but they do seem to have been fairly useful too.
John Quiggin 06.23.13 at 8:16 pm
@Stephen If you read the comments threads on the previous versions of this post, there’s a lot of discussion of these points, which (on my reading) confirms the argument of the post.
Cranky Observer 06.23.13 at 9:00 pm
Despite hundreds of thousands of words published and spoken by the man, two soul-baring autobiographical books, and the brutal vetting of two Presidential campaigns even those who have worked extensively in the higher reaches of politics and have worked with him personally can’t agree on what Barack Obama’s intentions are. That’s for a major actor who has been under a public microscope most of his adult life. How the heck can our intelligence agencies, which are not staffed by demigods, discern the intentions of players they don’t know from cultures they typically don’t have much understanding of?
Cranky
novakant 06.23.13 at 9:10 pm
But economic espionage – which hasn’t been discussed much , but is certainly happening – could be hugely profitable.
You say that as if it was a good thing.
Tim Wilkinson 06.23.13 at 9:38 pm
Cranky Observer – did you intend to agree with me?
Marc 06.23.13 at 9:43 pm
Assume that everyone else except you employs spies. Why would this end up being to your advantage? A world in which, say, all communications of your citizens are known to foreign governments, but not vice versa, seems to fail some pretty obvious tests of self-interest.
Cranky Observer 06.23.13 at 9:44 pm
Tim,
I’ve just managed to confuse myself by completely misattributing the quote I used in my 9:00 pm comment. In any case, my example of the difficulty of understanding the intentions of a famous political figure in the US was intended to reinforce the point about the difficulty in discerning intentions – which I think was your point. I apologize for the incorrect cite of your text.
Cranky
Bruce Wilder 06.23.13 at 9:55 pm
Applying a game theoretical framework to the interactions of diplomatic anarchy, in a loose sort of way, it seems to me that one thing, which leaps out immediately is that an individual Player, in equipping herself with strategic options, would want, with regard to her own communications, a potentially quite elaborate gradation, between public, private and secret.
Imagine these as layers, simply to make it easier to name and rank them.
First, on “top” so to speak, are the completely public declarations, the “broadcast” layer, which are communicated in ways meant to reach a lot of people not directly employed in her own organization. Maybe, there are several sub-layers to the broadcast layer, with varying degree of intended reach, but all public.
First story below the general broadcast level are the communications, broadcast to members of her organization. Physically or logically open (“transparent”), these are communications within the Fishbowl.
Below the Broadcast and Fishbowl layers, are the Private communications — the communications, which, as they say, might prove “embarrassing”. Polite behavior requires white lies and keeping inconvenient truth to one’s self on occasion. Tact. Private information for the firm might include the salaries paid to individual staffers and middle managers; for the householder in the neighborhood, it might be how much the household earns in income, or paid for the new car (is it leased?), or what medications the husband takes.
Private information isn’t really secret — and may be plausibly guessed — but privacy is potentially important to maintaining a degree of strategic autonomy. One bargains over particular disclosures and authentications — like estimating income for the bank in applying for a loan. The general attitude is: “People don’t need to know my business.”
For a state, within which — that is, within the body politic — there’s a lot of strategic behavior contesting control of the state’s behavior, the analogue to “private” communications for individuals and families, is a bit vague and confused.
Certainly, the conventions between states honors norms of privacy, even if it honors them in the breach. Plausible deniability is a treasured strategic option, even when the “plausibility” part applies only to Andrew F.
Private isn’t really secret, but it is basic to certain ritualized exchanges of disclosure, declarations of commitment and the like. Disclosure of private information can threaten strategic autonomy.
Political scientists may have a term for what I am calling “private” as distinct from “secret”; I would have no way of knowing. In the U.S. government, the term of art is “classified”. Classified information is private to the government as an operating organization; classification determines who within the government (including its privatized operating arms) can be privy to the information, and seeks to prevent unauthorized disclosure.
I’m trying to think of some clever way to distinguish between “private” and “secret”. Secrets are facts our Player simply does not want anyone to know, not facts that she wants to deny or disclose, when convenient. It is hard to see the boundary between “private” and “secret” as anything but very wide, marked in many shaded gradations.
In the operations of the U.S. government, genuine secrets are not classified at all, because to classify them is to disclose them, at least within the circle defined by the classification and to record or document them in ways that involve possible uncontrolled exposure.
Cranky Observer 06.23.13 at 10:01 pm
If the damage their spying efforts do to their own society [1] exceeds the benefits that the nominal “enemies” receive from the spying, yes. This has been discussed theoretically and fictionally [2]. Real-world examples are hard to find since after WWI most of the world’s nations embraced the national security state concept with gusto, but modern Germany’s relative modest expenditures on such efforts and the performance of their productive economy compared to the US and UK’s might be a good place to start thinking.
I’d also be curious to know see an enumeration of the benefits of all this spying. Hiring Google to help with the police work of tracking down Osama bin Laden is probably a good use of government funds and moral capital. Amassing gigantic dossiers on every computer user on the planet? I’ll really need to see a list of the positive value of that activity.
More snarkily, those who have been around 10-11 year-old boys know that they will occasionally undertake activities such as determining who can slap their own face the hardest and absorb the most slaps before quitting. There certainly is going to a leader and a winner in that activity, but it seems to most adults that the only winning move is not to play.
Cranky
[1] A short list that comes immediately to mind, but by on means exhaustive, includes loss of trust throughout society, loss of espied industrial secrets hoovered up for ‘security’ reasons and either traded away for spycraft or simply leaked, diversion of monies needed for health care, schooling, etc to spying and “security” activities, destruction of a campaign to vaccinate against a deadly disease by allowing spies to use it for their purposes, and general overall deadweight loss resulting from over allocation of resources to “security” and “defense” establishments.
[2] Poul Anderson and (perhaps oddly) Tom Clancy both wrote variations on this possibility into some of their work.
Bruce Wilder 06.23.13 at 10:38 pm
I’m not much impressed by people, who think the recent disclosures by Snowden are an occasion for moralizing about the 4th amendment and having a ritualized outrage attack. There’s been an almost total technological disruption of our established conventions and norms of privacy over the last decade; this is just one aspect of that disruption, and not a singular breach, which can be remedied by untethered moralism.
That said, I do think what is going on in government is extremely serious. Not because I’m acutely worried that the government might seek a peak at emails, which Mr. Google and his friends already read. I’m annoyed that my email is so public. But, what worries me about the programs Snowden confirmed the existence of, is not what the programs do, per se, but that the programs, themselves, were (and remain, officially) secret.
The other thing that worries me is that the government has vastly increased the scope of communications and documentation, which it classifies, and is trying to ramp up the normative seriousness of breaching classification, with increased penalties and vigilance.
That the government is looking at my email network (like Linked In, Google Plus, Facebook, and who knows who else isn’t) or my phone bill isn’t the focus of my most serious political concerns. I think it would be better for my autonomy if my phone bill and email were more “private” in the sense of being more under my discretionary control to disclose, but my concern about the government’s breach of the 4th amendment is secondary. If my bedroom no longer had walls, I’d have bigger worries than whether the LAPD now thought it didn’t need a search warrant.
I am concerned that democracy, itself, is under assault, because so much information is classified that we no longer have the means to be an informed citizenry. Secrets and lies are two sides of the same counterfeit coin, when citizens have to “trust” official statements, which cannot be verified, the verification of which is made a violation of the Espionage Act.
The Transpacific Partnership “trade” agreement is so secret that it cannot be debated in Congress — who knows if they’ll be allowed even a ceremonial vote. The routing of pipelines, like the proposed Keystone XL, are now routinely classified, and potential protesters are identified by corporations to law enforcement as “terrorists”.
I think it is crazy behavior that the Obama Administration is exhibiting, but I also see that behind it, is a panicked reaction to the technological disruption, which has made access and processing of information so easy, so nearly costless — the same revolution, which has exposed everyone’s personal information to scrutiny. A generation ago, the intelligence agencies devoted significant resources just to mapping territory and photographing buildings, and, now, everyone has access to better maps and aerial photography than the CIA had in 1970, and they have it for almost everywhere. My city and county have put on line vast amounts of city planning data — things I would not have guessed even existed: I have a map of landslide zones in my hilly neighborhood; one sweeps the street in front of my door; who knew?
A lot of things people in authority never really worried about, because the resource costs of gathering and systematically verifying the information made them either practically unknowable or the effective monopoly of the sovereign government, well, now, all of a sudden, they do have to worry, they do have to think through the implications, and they aren’t necessarily very good at thinking through implications. And, they are getting seriously, dangerously carried away with themselves, and being incentived (horrifying word! Glad Google’s spell check doesn’t like it) to go overboard, by our accelerating corruption and the demand that tax dollars be funneled to for-profit corporations for makework.
John Quiggin 06.23.13 at 11:04 pm
Replying further to Marc. If other people are spying on you, it makes sense to adopt a range of countermeasures – everyone is worse off than if no one is spying, but you are better off than if you allow the other side free access.But if you spy in return, and the other side adopts its own countermeasures, that only increases the damage for everyone.
Even the usual deterrrence arguments don’t apply here, since retaliatory espionage is secret. The best deterrent would be to severely punish foreign spies, close down embassies that harbour spies, and make it clear that any unauthorised spies on your own side will be left to their fate.
William Timberman 06.23.13 at 11:27 pm
I didn’t think I’d be able to find this quote, but for today at least, Google was my friend. From Eliot, just as I remembered — it seems that my memory was also my friend today, which in recent years has rarely been the case. Very like Eliot it is, too — the Eliot who was deeply skeptical about modernity.
Frankly, these questions don’t trouble me overmuch. The wisdom and knowledge we’re looking for are still lying where we left them, despite the odd mix of hubris and hysteria than modernity has placed between us and them. All we have to do is pick them up, dust them off and apply them to our new situation. This isn’t likely to be an easy task, but then it never was.
How is this relevant to secrets, lies, and the efficacy of intelligence gathering? Well, for starters….
SIGINT is for lazy, paranoid people, people with more money than sense; HUMINT better acquaints us with the rest of the world’s intentions, but also acquaints us with its fears and aspirations as well, an understanding of which might alter our allegiances, and create a breeding ground for treason.
Can’t have that.
john c. halasz 06.23.13 at 11:32 pm
http://www.mcclatchydc.com/2013/06/20/194513/obamas-crackdown-views-leaks-as.html#.UceFc4Vpe_U
Suzanne 06.23.13 at 11:33 pm
@20: I’m not much impressed by people, who think the recent disclosures by Snowden are an occasion for moralizing about the 4th amendment and having a ritualized outrage attack.
I’d say a little 4th Amendment moralizing is perfectly timely. And people are not necessarily having “ritualized outrage attacks”; some of them are plain outraged.
In fact here in the U.S. there seems to be relatively little in the way of impassioned outcry, rather the opposite.
I believe the favored neologism is “incentivized” rather than “incentived”, and spell checks seem to be just fine with the former, I’m sorry to say. They are both equally horrifying.
Rich Puchalsky 06.23.13 at 11:51 pm
“I’m not much impressed by people, who think the recent disclosures by Snowden are an occasion for moralizing about the 4th amendment and having a ritualized outrage attack.”
This is all wrong. Bruce’s comments are good for some things, but too much of them seem to be animated by a lost and possibly mythical era in which people created a competent government for good reasons.
Now that the disclosures are underway, how are we actually going to get change? Through public opinion? No. Public opinion makes the President go be interviewed by a tame journalist to deny everything, and it doesn’t matter how plausible it is. There are no levers of power by which the public can do anything directly. Protests are futile, the GOP is worse, etc etc.
So if there’s going to be any change other than via building the next generation of Internet tools outside the U.S. with heavy levels on encryptions, it has to be official change working through the levers of the U.S. government. Those paths come down to legislative, regulatory, or judicial. Legislative is not going to happen. Regulatory is going to be focussed on trying to make sure that low-level staff don’t listen to sex calls for laughs like they did to U.S. soldiers in Iraq. Judicial comes down, in the end, the 4th Amendment. It doesn’t matter whether you think it’s outdated, or whether you think the real issue is our privacy setup. There is no leverage there, so you can think whatever you want, and no one will care.
What’s going to matter is the 4th Amendment. And what’s more — the Supreme Court aren’t robots. They’re very conceited human beings, and they are by no means immune to public concern — not out of any desire to do right, but because they like being admired. If people want a better privacy regime, outrage about the 4th Amendment is a necessary part of the only path I can see towards getting one. Because the SC in the U.S. rules on interstate commerce too, and think they’re going to be markedly less sympathetic to businesses poking into everything if they ruled that government can’t.
Which is not to say that people’s outrage isn’t real. But real or not, there is at least a possible causative chain through which that outrage does something. What’s the chain otherwise?
Aaron 06.24.13 at 12:07 am
Even the presence of communication can be a piece of information, no matter how heavily encoded. This is the whole reason why the collection of metadata is an issue, regardless of whether the NSA actually reads emails. I am sure it is fairly difficult to have a conversation with a friend in Yemen without it being flagged as suspicious–which would force less efficient means of communication that might still be traceable.
You are also making some heroic assumptions about the ability of others to maintain coded communication at all times. Certainly foreign agents may be able to pull off, but what about foreign politicians, people under the influence of drugs (or narcotics dealers), etc. What is true for a perfectly rational agent probably isn’t feasible for someone who occasionally assumes they are safe. If there are mental costs to maintaining coded communication, then there are also likely mental lapses.
Until the advent of computer technology, surveillance of communications was not a huge part of intelligence, in part of the reasons you mention. “Flipping” insiders, bribing officials, establishing “who’s who” in opaque governments, and supporting dissidents were more effective. So there are good reasons to disband spies, especially domestic ones, but I don’t know that surveillance has ever been a prime expense.
Andrew F. 06.24.13 at 12:25 am
If other people are spying on you, it makes sense to adopt a range of countermeasures[…] severely punish foreign spies, close down embassies that harbour spies, and make it clear that any unauthorised spies on your own side will be left to their fate.
Countermeasures like that shouldn’t be necessary if the problem with spying is simply the possibility of disinformation. But a need to deter the activity of espionage even in peacetime implies that there may be value to it even given the possibility of disinformation.
Indeed, the implication is that intelligence collection that cannot be subject to the countermeasures you suggest, which would include signals intelligence, may be a wise course of action.
The real problem with the overall argument, though, is the implicit assertion that the odds of discerning information from disinformation are 50/50. Not entirely certain that the argument does contain that assertion – if not I’d welcome an expanded description of the premises.
Bruce Wilder 06.24.13 at 12:56 am
Discerning information from disinformation may be practically less of a concern than the possibility that almost everyone in your SIGINT office is an authoritarian follower, prone to paranoid hostility to the “other” and easy exploitation by sociopaths, with the consequence that opposed states, their
national homeland security apparatus hijacked by respective fascists, falls into a symbiotic relationship with their opponents, exaggerating and exacerbating the threat to boost their own budgets, steering the ships of state toward eventual collision.And, the smaller the actual threat, the less existential the threat, the less interest in state security policy there is, among the people, who are neither authoritarian followers or sociopathic dominators. If the threat from opposed states or factions was truly existential, there’d be lots of people, who would be interested in all the varieties of human intelligence and influence, involved in the politics of foreign policy. People, who would want to provide foreign aid for schools and clean water, send the Peace Corps, have cultural exchanges, fund the French Institute and the American University, etc.
Without a real existential threat from the outside, the only people left doing foreign policy are generals, and crazy ones, at that, who are rewarded for their incompetence in perpetuating insanely expensive wars.
Bruce Wilder 06.24.13 at 12:57 am
My strike went wrong. I only intended to strike national in favor of homeland (a phrase that nauseates me).
Bruce Wilder 06.24.13 at 1:03 am
Rich Puchalsky @ 25
You think our Federalist Society Judiciary is going to go all civil liberties for us, and, despite favoring corporate feudalism in every other context, they are going find some reason to “be markedly less sympathetic to businesses poking into everything if they ruled that government can’t”.
OK. I really have nothing to add. Wow.
Omega Centauri 06.24.13 at 1:58 am
John, I’m not convinced of the non-value of intelligence. Suppose we knew our opponent had a set of potentail moves available to him, say 100. Now our spying comes up with option X. By itself this is for the reasons you state of little to no value. However suppose I have some independent means of verification -but I couldn’t possibly afford the obvious shotgun approach -verify All possible moves. But I can select a small subset, and I would select X for verification.
And then of course the real world is not a game theory math problem. There are real nonzero probabilities that the opponent will do something not predicted by the game theory, -he may not be quite as rational as we presume. So what happens in the real world is only roughly approximated by game theory.
Rich Puchalsky 06.24.13 at 2:05 am
No one can be entirely sure why the SC does anything. Remember all the Kremlinology about why Roberts didn’t find Obamacare unconstitutional? It’s possible that, Federalist Society or not, some judges might have actual civil rights concerns while others just want to make Obama look bad. And the effect on business would only happen in some future SC.
The fact remains: the 4th Amendment is the only lever of power that has a chance of actually doing anything. Snowden’s revelations gave the ACLU standing to pursue part of this in court. Whether you think it’s likely or not, I see no other route of action.
John Quiggin 06.24.13 at 2:56 am
“So what happens in the real world is only roughly approximated by game theory.”
Sure, but why suppose that irrational behavior on either side (in addition to the original, irrational, decision to play the game at all) is going to improve outcomes. More likely, it will mean even larger losses for both sides.
John Quiggin 06.24.13 at 3:08 am
“I’m not much impressed by people, who think the recent disclosures by Snowden are an occasion for moralizing about the 4th amendment and having a ritualized outrage attack.’
Opening statements like this lead me to ignore everything that follows. Maybe that’s just me, but if you are commenting on my posts, you might want to bear it in mind.
Anarcissie 06.24.13 at 3:08 am
If we’re talking about a government or other powerful body spying on ordinary people, I think the primary profit is in the terror created, and then there is a secondary profit in creating a large, bureaucratic industry with a big budget and many posts to fill.
Cranky Observer 06.24.13 at 3:39 am
What exactly are these “moves” going to be? Against the United States, that is. Which power threatens to take a drink of water from the Ohio River? Examples of how useful spying can be are taken from a period of time when the US was locked in a potentially existential conflict with arguably the most accomplished military power in human history – and 3 of the 5 years of that conflict were a Congressionally-declared war. Outside of periods of declared war, where is the evidence that spying has net value to the citizenry? Or we don’t know because we can’t handle the truth?
Certainly there are entities that use terrorism, which might attack the United States, and which need good solid police work to be kept in check. Including, if necessary, judicially-reviewed electronic spying. Other than that? Is the PRC really going to start an overt conflict with the US? If we’re concerned about the PRC subverting us (the US, in my case) economically, perhaps we could spend the spy money on directly subsidizing our indigenous consumer manufacturing industry and getting us the heck off our addiction to oil rather than pouring it down a game theoretic rat hole? Perhaps that might make us, I dunno, safer? Neither the Treasury nor the Justice Dept has enough money to hire lawyers to investigate the events leading up to the 2008 financial crash, but we DO have the resources to track down FDA scientists “leaking” information about laboratory fraud to Congress? This allocation of resources makes us safer?
In our discussion we could at least give a nod to the alternative hypothesis that spying has zero or negative value and not simply run with the unstated premise that spying has some positive value that must be balanced. Cost/benefit analysis doesn’t work too well when you fail to analyze the case where benefit is negative. Andrew F is no doubt unwilling to even consider that such a hypothesis can even be constructed without violating some natsec version of Godel’s incompleteness theorem, but I like to think of CTers as being a bit more creative in their thinking.
Cranky
Lee A. Arnold 06.24.13 at 3:53 am
Well the intelligence agencies would not usually describe their successes publicly. It would tip off other malefactors to use different means. But I think that listening in to everybody is less useful than might have been imagined: (1) Any serious terrorist is not going to use electronic communication. (2) Breaking codes is only useful if you have lots of messages over a limited universe of possible referents (e.g. you are looking at the Nazi high command + the u-boats, for example). (3) Since we are dealing instead with an infinite universe of all possible violent acts, the only useful information is (a) messages from stupid teenagers expressing their wishes to kill everybody at school, and (b) meta-data of phone calls of suspects who were detected in other ways, such as by “humint”.
I am not a big fan of game theory but I think it requires the hand to be played: so it makes sense in terms of how to convince the Soviets that both sides should reduce their nuclear arsenals. But I don’t see how it would be effective for reduction of spying. Spying does not (so far) lead to Armageddon, or the end of all games. And spy reductions would never be open to verification by the other side(s).
As for the days of the gentlemen? They passed by with Renoir’s Grand Illusion, and The Life and Death of Colonel Blimp. They might come back again, but it will be after cessation of hostilities everywhere, and “the dawn of peace, pure, perfect, and perpetual, waking the weary of the world.”
Matt 06.24.13 at 4:54 am
I don’t see why “serious terrorists” wouldn’t use electronic communications. Snowden confirmed that the NSA has no way to attack modern cryptography head-on, and must instead look for ways to sidestep it. I do remember reading news reports a while ago about terror cells that trusted to their homebrew code systems instead of PGP or the like. They “cleverly” shared a single email login to a web mail account instead of actually send mail across the internet, but they still used (IIRC) Yahoo mail instead of something out of Western jurisdiction. Had they stuck to modern cryptography, the message contents would still be awaiting a breakthrough in computer science.
TOR prevents traffic analysis, but the exit nodes (perhaps operated by the NSA itself!) can snoop on unencrypted content. Encryption prevents content analysis but it’s still vulnerable to traffic analysis if you just send encrypted email. Both must be combined if you’re really serious about security: use Off the Record IM over TOR instead of PGP email, or share encrypted messages through TOR via pastebin or another site that doesn’t require registration.
This is a lot of work if you aren’t trying to protect the contents and patterns of your communications. The low-hanging fruit is for everyone to enable encryption as a matter of course for internet services, so the contents (if not patterns) of your communications are safe from data dragnets. Again, I urge whoever runs the infrastructure for Crooked Timber to look in to getting a dedicated IP address and a self-signed SSL certificate.
PS This might be a duplicate post. Previous one was attempted through TOR and ironically appeared to fail.
Peter T 06.24.13 at 7:39 am
Maybe I am missing something, but would not JQ’s game-theoretic reasoning apply to any situation where two parties have some opposing interests and limited ability to keep all their interactions with other parties private? In which case we can see that posting diplomats to talk to people is a waste of money (they will just lie and be lied to), gathering information is ditto a waste of money (it will just be either useless or misleading) – in fact any discussion at all is a waste of time. We could save on more than the intelligence budget.
reason 06.24.13 at 8:31 am
Marc @16
Makes a very good argument, the answer to which is that it may not be of advantage to have spies, but it is of advantage to appear to have spies.
Nice paradox.
reason 06.24.13 at 8:32 am
Consumatopia @4
Yes, I think this is a valid point. Government vs government is one thing, government versus small isolated terrorist cell, quite another.
ajay 06.24.13 at 8:49 am
The main point I want to make is that it’s time for Britain to get out of the spy game. More than any other democratic country, Britain is addicted to spies and their natural counterpart, Official Secrets. From Burgess and McLean to the present day, the spies have been a constant cause of embarrassment and worse. On the other hand, there’s no evidence that they’ve ever found out anything that was both useful and sufficiently reliable to act on (in this context, I’m excluding wartime codebreaking, which is always useful since, at a minimum, it disrupts enemy communications).
Since this was originally written, there have been many court cases in Britain which have made it clear that the spies – in this context the Security Service and Special Branch – have indeed found out lots of things that were useful and reliable in the context of counterterrorism. There’s also been a lot of information about the role of good intelligence in bringing the Northern Irish terror campaign to an end – by the late stages of the Troubles, the PIRA was extensively penetrated by the security services.
But it’s possible that JQ thinks that actually knowing anything about intelligence work makes you less rather than more likely to come to the right conclusions about its utility.
Harald K 06.24.13 at 9:54 am
A shorter way to make the point, without direct reference to game theory: Information gathering based on betrayal of trust can’t be sustained. Sooner or later, if you allow yourself to act on the information at all, that trust is going to dry up.
Considering that trust is of grave economic importance to Silicon Valley, the US government should not want to reach that point.
Katherine 06.24.13 at 10:51 am
I’m having a flash back to a West Wing episode where an outgoing SC Justice excorates the President for driving straight to the middle of the road, a Presidential advisor declares that issues of privacy will be the next big civil rights thing, a SC Justice candidate is discounted because he doesn’t accept that a right to privacy is contained in the US Constitution, and in the end a new Justice is appointed who wants privacy for all. Hurrah for the Democrats!
Once again, fictional politics and politicians trump the real ones hands down.
Ronan(rf) 06.24.13 at 11:44 am
“There’s also been a lot of information about the role of good intelligence in bringing the Northern Irish terror campaign to an end – by the late stages of the Troubles, the PIRA was extensively penetrated by the security services”
Well there’s a lot of evidence that the security services did help to contain and wear down the PIRA, but there were also a lot of abuses committed by the security services where the impact was less positive. That was also never going to be the solution on its own, but a middle ground tactic until a political solution could be found, so there are questions to be asked about how effective it was as a strategy, were there alternatives ways of viewing and dealing with the conflict etc
Cian 06.24.13 at 11:59 am
Considering that trust is of grave economic importance to Silicon Valley, the US government should not want to reach that point.
You got your tenses wrong. ‘The US government should not have wanted to reach that point”. The Rubicon has been crossed – one more step in the surprisingly rapid decline of US power and influence.
Peter Erwin 06.24.13 at 12:40 pm
If other people are spying on you, it makes sense to adopt a range of countermeasures – everyone is worse off than if no one is spying, but you are better off than if you allow the other side free access.But if you spy in return, and the other side adopts its own countermeasures, that only increases the damage for everyone.
But this overlooks the utility of espionage for counter-espionage efforts. Getting access to the agencies that are spying on you can be extremely useful (as Aldrich Ames was for the Soviets).
You can’t always cleanly separate foreign espionage and domestic counter-espionage, even if it makes your game-theory models simpler if you pretend that’s how it works.
John Quiggin 06.24.13 at 1:07 pm
ajay, I have little direct knowledge of the UK Special Branch. On the other hand, I have plenty of knowledge of Australian secret police with the same name, formed on the UK model, and everything I’ve written applies to them. Although Australia doesn’t have a great record on civil liberties, we have at least managed to abolish these pieces of the police state apparatus.
Based on our previous discussions, your endorsement of the UK version is good evidence for the view that they are just as bad as their Oz counterparts.
John Quiggin 06.24.13 at 1:28 pm
And to amplify Ronan’s point, the abuses of which the various “special” forces (B Specials, Special Patrol Group, Special Branch) were a symbol did at least as much to prolong the Troubles as to resolve them.
bexley 06.24.13 at 1:32 pm
My second hand understanding is that Scalia is actually consistently decent on the 4th amendment.
John Quiggin 06.24.13 at 1:35 pm
@Peter Where was the Soviet espionage in the Ames case? AFAICT, Ames passed nothing to the Soviets other than info on US espionage efforts. How could he have done, since he was a full-time member of the espionage apparatus? The separation between espionage (Ames work with the CIA) and counter-espionage (what the Soviets did with his assistance) seems as sharp as you could ask. Of course, the Soviets had their own spy efforts, but a government with open commitments to refrain from espionage on their own behalf and to reward foreign spies who betrayed their colleagues could have done the same in this case.
Dr. Hilarius 06.24.13 at 2:05 pm
I share Bruce’s concern that classification of governmental actions now exceeds anything necessary for actual national security concerns. I know, I know, how to define “actual national security concerns.” Obviously, the possibility of a dirty bomb being planned qualifies but so many of the terrorist plots the government claims to have foiled have turned out to be nitwits encouraged to action, from mere talk, by the FBI. The Liberty City case comes to mind immediately.
It appears, however, that classification now exists to carry out policy without need for debate or approval by us unwashed citizens. Dissent is equated with terrorism. The 82-year-old nun and two other activists facing up to 35 years in prison for walking onto the Oak Ridge nuclear facility is a case in point. A purely symbolic protest is charged as “intending to injure, interfere with, or obstruct the national defense of the United States and willful damage of national security premises in violation of 18 US Code 2155.” Their trespass embarrassingly showed the lack of facility security, which may be their real crime. Animal rights and environmental activists, some admittedly causing property damage, are equated with al-queda.
Economic and trade policies are being cloaked in secrecy in order to implement them without public input. The Trans-Pacific Strategic Economic Partnership talks, cited by Bruce above, is a great example. The official line is that the talks need to remain secret or our potential partners won’t feel comfortable negotiating. The only party excluded is the US public. Support for this variety of secret is one of the few examples of bi-paritsanship in the US today.
Governments have always wanted to work without the annoyance of debate and dissent mucking up things. The War on Terror has been a godsend enabling anything to be covered by the cloak of national security. And without any chance of an inconvenient peace erupting.
bexley 06.24.13 at 2:51 pm
And as a bonus, the spies of Special Branch tried to smear anti-racism campaigners like the Lawrences.
Although if they really want to up their game they need to get lessons from the French on blowing up boats owned by enviromental groups.
Ralph Hitchens 06.24.13 at 3:09 pm
Plenty of balanced commentary above; Stephen’s is a good example. Intelligence collection ought to be separated by component when evaluating its effectiveness. SIGINT had good documented results in World War II and sporadically so during the Cold War — e.g., we developed a pretty good understanding of how the USSR & Warsaw Pact would organize for combat and execute war plans if the “balloon went up.” As for human intelligence (HUMINT) I would argue that from the Soviet perspective both Ames and Hanssen were effective from a counterintelligence standpoint, and the Walkers were fantastically effective in enabling serious improvement in Soviet submarine capabilities. That the Cold War ended with a Soviet collapse made all of this moot, but it should be considered in our discussion.
While the impact of HUMINT is difficult to assess in the GWOT, what SIGINT has done deserves mention. All terrorist organizations, from the 19th century anarchists to the Weather Underground to al-Qaeda devote most of their energy — 90% or more, I’d guess — to operational security. Al-Qaeda plays a “long game” because it’s the only game they CAN play with any assurance of security, in an age of (perceived) ubiquitous electronic surveillance. Human couriers handling key links really slows things down, and centrally-coordinated “real time” operations are largely impossible. This in turn means that key terrorist personnel involved in proposed operations must remain below the radar for longer periods of time, which is highly stressful and difficult, and plays to some of our advantages. The most egregious “what if/missed opportunity” of recent decades is probably the well-documented fact that in early 2000 CIA obtained the names of two al-Qaeda operatives subsequently involved in the 9/11 hijackings, and discovered that these individuals had multiple-entry US visas. Had they shared this information with the FBI the two might have been tracked upon entering the USA later in 2000, and their phone contact with several other hijackers monitored. The entire operation could well have been rolled up in advance, but for the CIA’s dreadful policy of “need to know.” All too often we are our own worst enemy.
Katherine 06.24.13 at 3:37 pm
http://www.guardian.co.uk/uk/2013/jun/21/mclibel-leaflet-police-bob-lambert-mcdonalds
And hey look, one of the people that wrote the anti-McDonalds leaflet that was found to be libellous against them (but damages were nominal) was an police officer undercover at Greenpeace. What an incredibly intelligent use of police time.
Because everyone knows that London Greenpeace is a hot bed of radical terrorism-type something-or-other. Or perhaps just a hot bed, since he managed to deceive four women into having sexual relations with him, and abandoned a child he’d fathered whilst undercover, just to put the toerag-cherry on top of the deceptive-little-shit cake.
Josh G. 06.24.13 at 4:11 pm
Dr. Hilarius @ 52: “It appears, however, that classification now exists to carry out policy without need for debate or approval by us unwashed citizens. Dissent is equated with terrorism. ”
A case in point: the deputy director of the Tennessee Department of Environment and Conservation recently told a gathering of citizens that if they complained to the agency about poor water quality and the complaints turned out to be unfounded, “that can be considered under Homeland Security an act of terrorism”.
Josh G. 06.24.13 at 4:15 pm
Ralph Hitchens @ 54: ” SIGINT had good documented results in World War II and sporadically so during the Cold War — e.g., we developed a pretty good understanding of how the USSR & Warsaw Pact would organize for combat and execute war plans if the ‘balloon went up.’”
And what was the ultimate significance of this? Fortunately, neither the American nor the Soviet governments were stupid enough to blow up the world, and if they had, then neither side’s contingency plans would have amounted to a damn thing. As soon as thermonuclear weapons started being stockpiled on both sides, the notion of a “winner” in a war between the US and USSR became absurd on its face. Both sides would have lost, and humanity itself might well have become extinct. And no secret information was needed to figure any of this out.
Coulter 06.24.13 at 4:24 pm
Josh,
Both sides believed the other side did not agree with you. Both sides tried to figure out what they could be missing to allow the other side to believe they could fight and win WWIII. We came close in ’63 and ’83 … http://en.wikipedia.org/wiki/Able_Archer_83
Peter Erwin 06.24.13 at 4:52 pm
Further to Coulter @ 58:
This, from an article in Time on Soviet general (and American agent) Dimitri Polyakov:
hix 06.24.13 at 4:55 pm
Terrorists should be able to avoid getting caught by the equivalent of a google search into private data when they communicate electronic.
It is not that much of a stretch that such big data grabs are by large used to intimidate the non violent, not even radical ones who are just too far to the left or right from the current government. Im coming from a situation here in Bavaria, where the Verfassungsschutz is unable to catch Nazi murderers for a decade, but surveils all members of the left party.
The US is even more comedic when it comes to “al-qaeda terrorists”. There is no one fixed network of evil people that can be fought with guns and surveilance. Rather all the paranoid hate pushes more and more people over the edge. Either over the edge where they become violent themself, or over the edge where they do not work active against the extremists themself anymore. There would be far less people in Yemen that plan to kill Americans and those who do plan would bei reigned in by the locals if that place had just been left alone.
Andrew F. 06.24.13 at 5:13 pm
But according to the post the Soviets could not possibly determine whether Ames was giving information or disinformation. Therefore counterespionage in the form of obtaining confidential information from another country is as useless, according to the post, as espionage. Why would documents pilfered for counterespionage purposes be any more information rich than documents pilfered for other intelligence purposes?
Peter T @39 nicely demonstrates the problematic implications of the post’s argument.
Whether intelligence services are abusive of civil liberties is irrelevant to the question of whether they can gather useful information. The question raised wasn’t whether British intelligence enabled abuse, but whether it gathered and recognized as such useful information.
Cranky, intelligence is often very useful even in the absence of overt war. Knowing China’s intentions and capabilities with respect to Taiwan, for example, is as important to helping preserve Taiwan as it is to avoiding the outbreak of war. Obviously there are numerous other instances where intelligence would be useful. And let’s not forget that our track record of predicting wars well in advance is very bad, and that intelligence capabilities can require some time to develop.
Finally, I’ll just note that eavesdropping on enemy communications in war is not valuable primarily because it disrupts enemy communication. Indeed, great steps are usually undertaken to ensure that this eavesdropping does not disrupt enemy communication. This is because the real value is in learning what the enemy is communicating, despite the possibility of disinformation.
Now, deception certainly occurs, even when it’s not directed at the eavesdroppers. Rommel, for example, when communicating with German command, apparently would sometimes exaggerate his need for supplies by understating what he actually had on hand; the British were not the target of this deception, but initially they fell for it anyway. Nonetheless, I’d hazard a guess that most would conclude signals intelligence to have been an important capability during WW2.
Lee A. Arnold 06.24.13 at 5:37 pm
Peter Erwin #59 — Probably the same reason that the Israeli military leadership believes that the Iranian leaders are not crazy, either.
Ronan(rf) 06.24.13 at 6:29 pm
IIRC, poor intelligence was one of the main reasons that the Cold War developed as it did (both sides misjudging each-others capabilities, intent etc) Robert Jervis wrote a bit about intelligence failures and why he thinks they occur
http://home.aubg.bg/students/DRK090/Jervis.pdf
I guess there are arguments for and against ‘professionalising’ foreign policy, but personally I would be wary about thinking that these institutions and the people they advise are anything but barely competent. By extension I, personally, would be pretty wary about giving more power to barely competent, self absorbed small time tyrants.
Also, does anybody know to what extent the policing tactics used in the Troubles (informers, cameras etc) came to be used in Britain in general? Did these tactics grow out of the Troubles or were they always used?
Tim Wilkinson 06.24.13 at 7:30 pm
The spying arms race is parasitic on the secrecy arms race, of course, and the latter needs to be de-escalated. As others have noted, private corporations are part of the equation here, with their trade secrets and ‘commercial confidentiality’, which becomes part of the National Interest which is functionally equivalent to National Security.
The vast amount of unnecessary secrecy we’re faced with, which has to be protected from domestic populations if it’s to be kept secret from ‘the enemy’ – and often even when it’s not, note well – largely functions to serve greed for power and riches. It provides cover for some grotesquely undemocratic activities, sows mistrust and fuels paranoia. I’m not optimistic that official and corporate secrecy are going to be broken down, of course, but still.
Spying itself, like secrecy, is always going to be applied to the domestic population as well as to others – there’s a whole culture, and a thriving industry, of espionage which can never really be properly controlled except by being reduced across the board. If power corrupts, then essentially secretive and duplicitous forms of power do so all the more. And it’s insanely expensive – the NSA and CIA budgets are enormous – and the black – secret and illegal – budget of the CIA alone has been estimated at a $trillion p.a. – citation available via Google I expect (not that ‘legal’ means much in the context of the CIA which has always tended to lead rather than follow where legal authorisation is concerned).
Meanwhile, the Very Serious defences offered for the vast intelligence capabilities maintained by the US in particular are deeply unconvincing on inspection. Leaving aside the issue of the human and other resources used in maintaining a massive ‘security’/intel apparatus, I’d focus on a couple of directly relevant issues:
Tim Wilkinson 06.24.13 at 7:31 pm
…
1. Success stories
Many supposed successes of spying are misattributed: some bit of information was received which would, if believed, have prompted a certain course of action; that course was adopted; success! But in fact, as JQ’s model would predict, these bits of information necessarily trusted or understood except in retrospect – funnily enough this point is routinely accepted when intelligence failures are concerned – as in, for example, Pearl harbour.
In most of these putative cases, the decisions were taken not on the basis of espionage-based factoids but on other grounds. Someone mentioned Gordievsky, the double agent and subsequent defector with convert’s zeal, who was also a dedicated self-aggrandiser and routine flinger around of dubious accusations of spying against others. He claims to have saved the world from nuclear war by informing the Reagan administration that its bellicose brinksmanship was – unsurprisingly – causing genuine consternation in the Kremlin and risked war given the inevitable hair trigger for reaction or even attempted pre-emption.
But Andropov wasn’t believed when directly told the US this – they thought his warnings were partly for domestic consumpton and IIRC, partly a bluff to discourage the military buildup of Pershing missiles in Europe (why they supposed he would be worried about that is a mystery). And neither was Gordievsky considered credible, for exactly the reasons adumbrated by JQ. Since he’d been reporting Soviet fears of a US first strike since 1981, it would certainly seem that his reports weren’t being regarded as a decisive in this matter.
Perhaps most relevantly, the message that the USSR were worried about a first strike was also being relayed via ‘back channels’. This kind of unofficial diplomacy based on building trust in ongoing, fairly open and equal relationships is an ideal method of avoiding the kind of misunderstanding portrayed in the legend – far better than trying to assess the reports of self-serving, duplicitous chancers like Gordievsky.
But in any case it’s very unclear what positive developments Gordievsky is even supposed to be credited with, since none of this made much impression on the actions of the Reagan White House, which was determined to go ahead with military escalation aimed at bankrupting the Evil Empire. In 1983 Reagan was addressing the nation on TV to announce the Star Wars project, and warning them that ‘there will be risks’: not really the actions of someone who was unaware of, nor concerned about, the panic that his military escalation was causing in the USSR. He was also given to babbling about Armageddon and the ‘end times’.
This is fairly typical of legendary espionage successes. Reports based on spying are not reliable, so to be useful they either need to be independently corroborated, in which case the ‘corroborating’ evidence is generally doing all of the work, or to be purely ‘steganographic’ secrets – those which if suspected or even considered are ipso facto revealed. The easily investigated lead might be an example of this – or the plot so fiendish that it can’t be guessed at, yet at the same time so brilliantly simple that it seems obvious once revealed.
But for obvious reasons those whose business is keeping secrets prefer to have more protection for them than mere obscurity. Kerckhoffs’s principle, dating back to the late C19th, provides an example of this. Of course, using unreliable sources, hunches and so on to gather leads is much more appropriate where the adversary is an ordinary criminal, for example – they are much more likely to be relying on not being suspected, and they are also drawn from a large population of suspects. If one is more concerned with getting a conviction than with determining the guilty party, then of course this applies in spades. And the same of course applies to ‘policing’ non-criminal activities among the general public. Which leads nicely onto:
Shatterface 06.24.13 at 7:32 pm
The purpose of a panopticon isn’t to watch everything that people do but to make them act in a way they would do if they knew for a fact they were bein watched.
Whether or not spies ever discover useful information the possibility that you are being spied upon puts restraints on your actions. It also foments distrust between those who may be being spied upon and that makes organisation difficult.
By the way, according to Alex Butterworth’s fascinating book on 19th Century radicalism, The World That Never Was, Karl Marx’s inner circle was so thoroughly penetrated by the secret services that they actually examined his piles.
Tim Wilkinson 06.24.13 at 7:35 pm
…
2. Type I errors
The defenders of the standard Very Serious view are, predictably, prone to giving intel agencies the benefit of every doubt. It’s assumed that since we don’t know what they are up to, they must be having lots of successes which they simply can’t reveal. The agencies themselves have an obvious interest in reinforcing this impression. But why on earth should we believe this particular brand of ‘if you knew what we know’ assurance? After all, there is every reason to suppose that cet. par., failures are more likely to be covered up than successes. and the issue is complicated further by the fact that the kind of intel failure usually considered is what we can informally call the type II error, or false negative: cases in which espionage turned out to be ineffective (perhaps because there wasn’t enough of it, we might be led to suppose), but not actively damaging.
But when we consider Type I error – the false positive – espionage may be seen to have a negative value in certain cases. This might take the form either of directly acting on false reportsor of refraining from action on the same basis. This is a bit complicated, because decisions are generally taken on the basis of other kinds of information, with espionage being drawn on only to confirm existing prejudices. But that’s precisely because it’s so well recognised that espionage-based intelligence has very low real value compared to other sources such as public domain information or direct observation. And such cases are still, as in Iraq, a case of espionage having negative value from a naive National Interest point of view.
I suppose the German reponse to Operation Mincemeat would be a good example of intelligence with a negative value, though this was not strictly a case of espionage – again, precisely because the British realised that it wouldn’t be believed. But if the pro-spying contingent’s claims that spy-based info is acted on are accepted, then it must follow that in some cases such info has been wrong. Of course, where major actions have been taken on the basis of false information, one would expect this to be covered up pretty thoroughly and all involved to be given medals.
Known examples of genuine Type I error tend to be of the witchhunt kind – where the accuracy of the information is not really of much interest. This includes so-called ‘McCarthyist’ excesses in which we might include Angleton’s rounding up of innocent people accused of spying by the likes of Gordievsky. It also very probably includes a lot of GWOT cases – the Feds have observably been fitting up harmless losers (‘Kramer jihadists’) by means of agents provocateurs, and this overzealous approach to identifying ‘terrorists’ is clearly shared by the CIA, for example in the recursive method of torturing people until they hand over some names of other people who can be subjected to the same treatment. No doubt this unconcern about false positives extends to the use of paid informants and retained agents. And it’s the established practice of indulging in spy games that provides the capability and rationale for carrying out this kind of grim activity without even the discipline imposed by the need to get things past a judge and jury.
But of course that kind of thing is not Spy vs Spy, but spy vs. innocent citizen, random idiot or at worst, criminal.
John Quiggin 06.24.13 at 7:40 pm
I don’t usually bother with Andrew F’s trolling, but this is too good to miss. The Soviets didn’t have to worry about whether Ames was misinforming them – they could just arrest everyone he fingered on the basis that “God will know his own”. The effects, for the police state, would be beneficial, even if everyone arrested was innocent. And of course, once they had the suspects in custody, there would be no problem getting confessions of guilt, regardless of the facts.
For further info on this topic, see Guantanamo Bay.
John Quiggin 06.24.13 at 7:42 pm
I see Tim W @65 has made the same point in his final para
Tim Wilkinson 06.24.13 at 7:48 pm
And expatiated a bit more on it in the 3rd instalment, currently in the spam trap.
William Timberman 06.24.13 at 8:33 pm
I’d also like to point out yet again that if we allow the spooks and their defenders to define reality in their terms — us vs. the terrorists who hate our freedoms, etc. — then of course they’ll always be right about everything. The rest of us can then sit around explaining to one another how really, really effective intelligence is supposed to work, and agreeing that if the intelligence we’ve got doesn’t work as well as it should, there are all sorts of clever ways to make it work better, completely overlooking the fact that our continued participation, perhaps even our continued existence in the world we were born into depends on the sufferance of those we’ve already agreed have higher knowledge of its workings. The reality that President Obama is defending, and Andrew F. is so patiently explaining to us here, is in fact a kind of madness, and as such its premises have no legitimate claim an anyone who still has any sanity left.
Dr. Hilarius 06.24.13 at 9:25 pm
Looking at my bookcase, McCoy’s “Politics of Heroin in Southeast Asia” caught my eye. The CIA’s involvement in facilitating the opium trade and the “secret” war in Laos were secrets only to the American public at the time. In Thailand and Laos the activities were known, they just weren’t publicly acknowledged. Once again, only the US public was in the dark, darkness assisted by new organizations which declined to investigate or publish these open secrets.
Media complicity has been and remains (increasingly in my opinion) a large factor in US citizens being ignorant of what happens in their name. Unauthorized leaks are ever more vital in a time when David Brooks and Tom Friedman are regarded as foreign policy experts. Can’t recall the name of the allegedly ex-CIA spook who gives TV interviews defending torture and telling us how grateful we should all be for guys like him, keeping all that unpleasantness from us and keeping us safe. As clear a psychopath as I’ve ever seen.
John Quiggin 06.24.13 at 9:58 pm
It would be worth reviewing the Iraq WMD intelligence in the light of this discussion. In particular, in the provision of obviously false intelligence, who was in on the joke?
Blair, Bush and Cheney must have known, at some level, that they were being told what they wanted to hear. Ideally, as rational agents, they would have wanted to get this information through official secret channels, to be fed to (I assume) honest suckers like Powell, the NY Times etc, while getting the real facts through a separate channel so that they could get their story straight when no weapons were found.
But I suspect they weren’t consciously duplicitous enough to realise this. Even if they were, the difficulties of keeping the channels separate, without admitting their existence, are probably too great to be overcome.
Peter Erwin 06.24.13 at 10:56 pm
@John,
Stationing intelligence agents in foreign capitals, with the authority and resources to recruit and pay local citizens working for intelligence and police agencies isn’t espionage? Who knew? And it’s OK as long as they’re careful only to recruit people engaged in actual espionage, rather than counter-espionage for their own country?
Aldrich Ames was at one point promoted to head of the CIA’s Soviet counterintelligence branch — should the Soviets have backed away from him once they knew this? (It’s like you’re suggesting that the CIA had no interest at all in identifying Soviet agents
And of course Robert Hanssen worked on counter-intelligence for the FBI, and would thus be off-limits in your “gentleman’s rules” approach — and yet he was if anything even more effective for the KGB’s own counter-intelligence work than Ames.
Dimitri Polyakov, the Soviet general whose role as a CIA agent was betrayed by both Hanssen and Ames, provided the US with the names of several American military personnel working for the Soviets (and confirmation that Briton Frank Bossard was a Soviet spy), in addition to the wide range of Soviet policy and military-technology information he passed to the US.
Andrew F. 06.24.13 at 11:43 pm
John, the problem for the Soviets posed by your argument in the OP is that Ames’s information may be a deception calculated to cause the Soviets harm should they believe it. Certainly they can simply eliminate or imprison everyone implicated by Ames, but if the disinformation problem is as serious as you say in the post, then in doing so they have even odds of killing or imprisoning just the people that the US wants them to. Do you see the problem?
And if there are methods that the Soviets can use to ascertain whether the Ames source is passing disinformation, then presumably those same methods can be used to ascertain whether other sources of intelligence, for purposes other than counter-espionage, are passing disinformation. At that point we can no longer easily conclude that espionage is an irrational decision.
Tim, I don’t think many would claim Gordievsky, by himself, stopped the US from engaging in more provocative acts and thereby prevented a nuclear war, and so I do not think such a claim makes for a good example of a “typical” claim of intelligence success. Insofar as he is an intelligence success, he did furnish, as far as I know, as impressively large volume of information, much of which apparently proved accurate, over the course of his years of service as a British agent and KGB colonel. Nor do I believe that the US and UK were as likely to be fooled by any disinformation provided as they were to believe good information, which is what the post requires for its argument to actually work. I suspect you don’t either.
Ronan(rf) 06.25.13 at 12:01 am
When looking for something else I found this game theoretic take on national security secrecy, that might be of interest to someone
http://www.princeton.edu/~jns/publications/Is%20This%20Paper%20Dangerous.pdf
Peter T 06.25.13 at 12:47 am
Without in any way defending the practices of the US security state, let me say that JQ’s reasoning is spectacularly wrong. Intelligence collection does not target your communications with your ‘enemies”; it targets communications with oneself. These cannot be made in code without compromising one’s ability to function – they can only be secured. The spy is not asking “what do you think”; he/she is asking “what did you tell x to do”.
Illustrative intercept: Criminal 1: “The T-shirts have arrived!”. Criminal 2: “What T-shirts? I don’t want no T-shirts!” Criminal 1: “The white T-shirts. You know, the WHITE T-shirts.” Criminal 2: “Still don’t want no T-shirts.” Criminal 1: “Forget the T-shirts. We are talking about white stuff.” Criminal 2: “Don’t want no #@! white stuff either.” Criminal 1: “Oh for fuck’s sake! The heroin’s here.”
John Quiggin 06.25.13 at 1:58 am
Peter T, I have no idea what you think you are responding to here. When did I say that intelligence collection targets communication with your “enemies”? And I specifically made the point that, in a zero-sum conflict, monitoring enemy communications is valuable even if the only effect is to impose costs
Peter Erwin, I’ll assume good faith but I’m not interested in the kind of quibbles you’re raising here.
Andrew F, I’ve had my fun, and now it’s back to DNFTT as far as you’re concerned
Z 06.25.13 at 9:14 am
[T]here’s no evidence that [spies] ever found out anything that was both useful and sufficiently reliable to act on
With the caveat (explicit in JQ’s post) that we are talking about peace time, I came back from employment in the military intelligence exactly convinced by this fact. The reason is quite simple: intelligence is supposed to be research, but it is research done with the absolute guarantee that it will never be reviewed (because of secrecy issues). I think there are many more academics reading CT than member of the intelligence community, so that should be meaningful to them: what do you think would be the quality of research done in your field if you knew it would never ever be critically examined? Well that’s the level you should expect from intelligence agency: broadly correct, probably, and exceptionally brilliant but generally marred with sloppiness, wishful thinking and subtle errors.
So in a sense, I had a simple recommendation for improving military intelligence: make your findings public. Wikileaks as the ultimate intelligence service… Go figure.
Even if they were, the difficulties of keeping the channels separate, without admitting their existence, are probably too great to be overcome.
The separate channels existed whether they wanted it or not. For instance, the US Air Force intelligence agency (you know, the guys whose job it was to monitor Irak’s militay capability) took a major dissent with respect to the existence of operational WMD. In fact, they had to since their reports had repeatedly claimed that no such things existed in the past 6 years (as I know first hand, having collaborated to one of them). So one has to conclude that Cheney et al. deliberately chose their preferred channels, as was probably obvious anyway.
Andrew F. 06.25.13 at 2:07 pm
:) You’ve called me a troll twice now, where I’ve done nothing but take your argument seriously and express doubts that seem to be shared by quite a few others.
I think speculation, trotting out untested ideas, and in general bull sessions on blogs can be fun and useful. The criticisms and questions of the ideas put forward are a part of it, though. If you find such criticisms and questions to be trolling, well, then perhaps you intend for something other than discussion to result from the post, though I don’t know what.
The post is similar in some ways to “Who Needs a Navy?”.
You introduced a rather stunning claim – that because intelligence we collect can always be disinformation, over the long run intelligence collection and analysis is of no use to us. This seems really, really implausible – and you yourself have several times throughout the thread seemed to accept that in some cases, such as Ames, or Ultra, we actually have good odds of distinguishing good information from disinformation.
Like your claims about the utility of navies, your claim about intelligence requires more in the way of supporting facts – here there is little more than the postulation of a certain type of game. It requires some defense of why the assumptions underlying the claims should be accepted as true. To the end of examining those assumptions, various persons have brought up cases like Ultra and Magic, and Ames.
But you’ve ignored the point of those cases – that they provided good intelligence – and dismissed them with rather odd additional claims. You seem to think the utility of programs like Ultra and Magic in war is that they suppress enemy communications – when in fact their utility is that they provide excellent intelligence which we can often distinguish from attempts at disinformation.
You seem to agree that assets like Ames can be excellent sources of intelligence for counterespionage, but then you seem oblivious to the fact that the “disinformation problem” is as applicable to Ames as any other source. When this was pointed out to you, your response was that the Soviets don’t have to care whether the information was good or bad, because they could simply use it as evidence to kill/imprison everyone implicated – unaware, apparently, that the Soviets might harm themselves by doing so.
Perhaps the point of your post, and the “Who Needs a Navy?” post, is simply to make a question viable for public discussion, in which case critical comments may not seem helpful at this stage. I cannot tell.
W/r/t “separate channels,” the State Department’s INR provided analysis that explicitly disagreed with some of the conclusions in the October 2002 NIE, and the views of other members of the Intelligence Community on various subjects (such as the Air Force view on Iraqi UAVs and the Energy Department view on aluminum tubes) were available. There’s a much longer discussion to be had about whether dissent was suppressed in less obvious ways – selection of some analysts over others, for instance – and there’s no doubt that some in the Bush Administration adopted a very harmfully hostile attitude towards dissenting views. But there were dissenting views, and those views seem to have been available – certain the Secretary of State would have been very familiar with the INR’s views.
jay 06.25.13 at 5:53 pm
Kirian Healy has a rather interesting post about metadata,
http://kieranhealy.org/blog/archives/2013/06/09/using-metadata-to-find-paul-revere/
What it suggests is that the PRISM/NSA/GCHQ/22OtherAlphabetSoupAllies sweeps, combined with a little computer number crunching, can establish highly probable links and connections.
For Al Quida and it’s clones, this won’t provide much in the way of Intel, because they use messengers and coded notes, not sat phones and cell phones, for fear of a missile homing in on the signal and operational awareness of Us capabilities.
The metatdata could, possibly identify future/wannabe “Jihadists”, but only while they are getting their jihad on while surfing the net, tweeting Mosque buddies. This probably explains some of the National Security State’s ability to find assorted losers and the mentally deficient, to entrap and sell to the public as a Victory in the War on Terror.
Where the metadata is much more useful, is in domestic spying. If you want to know the connections between Keystone Protesters and Enbridge Protesters, the metadata can tell you. If you want to find out who’s leaking Monsato dirt to Project Seed, the metadata can tell you. If you want to find out what Private Company is looking to launch an IPO, the metadata can tell you.
Stephen 06.25.13 at 7:33 pm
John Q@49: “the abuses of which the various “special†forces (B Specials, Special Patrol Group, Special Branch) were a symbol did at least as much to prolong the Troubles as to resolve them”. Oh dear, oh dear. John: do try to learn a little about these matters. All that these demons of your imagination have in common is the word Special. The B specials (NI reserve police force, called out in times of emergency) did a good deal to suppress earlier IRA bombing campaigns up to 1962, but were abolished in 1970. How they posthumously prolonged the Troubles in subsequent decades is for you to explain.
The Special Patrol Group were an exclusively London-based unit for combating serious public disorder and crime that could not be dealt with by local divisions. If you can explain how they prolonged the Northern Irish Troubles …
The London and Dublin Special Branches were units of the British and Irish police, dating from the 1880s and 1920s respectively, concentrating on Irish Republican efforts to explode, shoot or rob targets they thought suitable in Britain or (post-1923) southern Ireland. Again, how that prolonged the Troubles … surely that shortened them?
I say ditto to Ralph Hitchens@54, Peter Erwin@ 59 and Andrew F@61. If I am then accused of being a paid agent of infamous forces, I will bear that with what little fortitude (considering the accusors) is needed.
Ronan@63: “does anybody know to what extent the policing tactics used in the Troubles (informers, cameras etc) came to be used in Britain in general?” Informers, double agents inserted into subversive or treasonable groups, and surveillance of their communications (with the technology available at the time) in England go back to the 16th century if not earlier. See Charles Nichol’s excellent “The Reckoning” for these methods in as far as they apply to the case of Christopher Morley aka Kit Marlowe. Cameras, etc, I don’t know: probably not too long after they were invented.
Ronan may find this unacceptable: but if you were one of Elizabeth I’s ministers, faced with plots to assassinate Elizabeth, what would you have done?
Tim W@67: “German reponse to Operation Mincemeat would be a good example of intelligence with a negative value, though this was not strictly a case of espionage – again, precisely because the British realised that it wouldn’t be believed.” Dreadfully confused again here, I fear. The British hoped the deception operation would be believed, as indeed it was. Earlier posts maintaining that it might not have been believed, and so doesn’t really count as an example of successful intelligence, are even more seriously confused.
Ronan(rf) 06.25.13 at 7:45 pm
I would have told them it was a consequence of being powerful ; )
Stephen, have you read Tony Geraghty’s The Irish War? It’s been a while since I read it and remember it being quite speculative, but I remember him making the case that the tactics used by the security services during the Troubles began to be used in policing in Britain more generally. (Not the more extreme ones, of course) I’m wondering how accepted his argument is now
And he’s an ex squaddie and all the rest so a reasonable voice on such topics I’m sure
Ronan(rf) 06.25.13 at 8:09 pm
And Stephen, in fairness like, I don’t think it’s hugely controversial to claim the B Specials had some role not only in sustaining the Troubles but creating it. Obviously you wouldn’t put to much emphasis on one organisation, but the fact they were abolished in 70 is a bit of a cop out
More specifically, and this is also as far as I can remember as I haven’t read up on it in years and haven’t read the specialist lit etc, but in the name of God it was British ‘policy’, in part, to ‘sustain’ the Troubles, if by sustain you mean use the security services to contain the paramilitaries (b/c they couldn’t defeat them) and wait for a political solution. That’s not to make a judgement one way or other, but the idea that the security services helped ‘sustain’ the Troubles (and that a lot of the excesses did lead to support for the paramilitaries) isn’t particularly controversial
Consumatopia 06.25.13 at 9:38 pm
I’m not sure I buy the original thesis, but I’m really amazed that people think it can be assailed by citing instances in which intelligence services make the right call. If you get a piece of information, and you’re trying to decide whether its a ruse or genuine, well, hey, you might make the right guess. That’s true even if you flip a coin to make that guess.
I’m amazed that Tim W had to point this out @67, and further amazed that pointing it out has done no good. Maybe there aren’t as many Type I errors as he worries about, maybe they aren’t sufficiently harmful to negate the value of intelligence, but there are surely enough of them that proponents of intelligence agencies will have to do more than cite occasional successes.
So far as I can tell, the original premise is that disinformation is easy, intelligence is hard. “The basic lesson of game theory for a game of bluff like that of espionage is that, as long as it is possible for counterspies to generate misleading information most of the time, spies are useless even when their information happens to be correct”. So citing instances of successful deception would actually support John’s point.
Again, I’m in no position to defend that original point, I’m not sure I agree, but some of the attacks on it make zero sense.
Suzanne 06.25.13 at 9:55 pm
@82: “… but if you were one of Elizabeth I’s ministers, faced with plots to assassinate Elizabeth, what would you have done?”
Funny you should mention Elizabeth’s security services. More than once positions taken by the Bush and Obama Administrations have put me in mind of Burleigh’s claim that Edmund Campion hadn’t really been tortured, because he could still walk.
Gloriana’s ministers would also have sent likely suspects over for a little enhanced interrogation chez Topcliffe and/or, per the Babington Plot, actually help facilitate the conspiracy they were monitoring (adding juicy bits to the correspondence to sway a reluctant monarch, when needful).
Tim Wilkinson 06.25.13 at 10:33 pm
Stephen – no, it is I who must ever-so-reluctantly point out ‘dreadful confusion’ on your part. Allow me to explain. perhaps the German response to Operation Mincemeat – i.e being fooled by it – would be a good example of intelligence with a negative value, though this was not strictly a case of espionage – again, precisely because the British realised that it – i.e. espionage – wouldn’t be believed. – which is why rather than conveying their misinformation via that route, they went to the trouble of doctoring a corpse etc.
I hope this resolves your confusion, and disabuses you of the notion that I thought the British decided to initiate an intricate plot that they ‘realised’ would fail.
While I’m at it, expanding on Consumatopia’s remark: Earlier posts maintaining that it might not have been believed, and so doesn’t really count as an example of successful intelligence? It wasn’t a case of successful intelligence. It was a disinformation operation. I assume you’re referring to sean matthews @3 who points out that the Germans didn’t simply believe the documents were genuine – the point, I take it, being that this wasn’t some exceptional case in which the usually (and usually accepted) reliability of intelligence was cunningly subverted, but simply a bad call made after the usual second-guessing had been made.
I admit I’m not an expert on Mincemeat (just as JQ admits to not being one on the details of exactly which organisations were involved in the tactics which he correctly points out ‘ did at least as much to prolong the Troubles as to resolve them’, thus rather disarming your attempted gotcha).
However, Wikipedia suggests that the decisive factor – which overrode the tendency of e.g., Goebbels to indulge in the usual convoluted second-guessing and eventually discount the source – was Hitler’s decision to accept it as veracious due to pre-existing prejudices (exactly the basis on which I’ve suggested untrustworthy intelligence tends to be trusted, converting the useless into the actively harmful.) peter @6, adds that the poor decision might also have been guided intentionally by fifth columnists in the Abwehr. I don’t see any confusion there.
jay 06.25.13 at 10:42 pm
Seems like some people think that guys like Bob Lambert and Mark Kennedy are hero’s.
Too bad the Force Research Group, the Glenanne Gang, and Ten Battalion UDR, along with other things, have slipped down the memory hole.
As http://www.thepolemicist.net/ points out, along with Verint and Narus, credit card records, phone logs and other NSA technologies outsourced to foreign companies, with software backdoors into the data.
Ronan(rf) 06.26.13 at 1:01 am
“I admit I’m not an expert on Mincemeat (just as JQ admits to not being one on the details of exactly which organisations were involved in the tactics which he correctly points out ‘ did at least as much to prolong the Troubles as to resolve them’, thus rather disarming your attempted gotcha). ”
Oh I don’t know if Stephens gotcha was so good even on its own terms. I was pretty sure the RUC had a Special control group which was heavily involved in collusion (and wiki seems to back that up) I’m open to be wrong on that though
Ronan(rf) 06.26.13 at 1:06 am
..patrol group
Ronan(rf) 06.26.13 at 1:15 am
Oh wow
http://www.tandfonline.com/doi/abs/10.1080/02684521003588070#.UcpAYYNOPIU
3 for 3. Bloody hell Stephen
Peter T 06.26.13 at 1:32 am
Mincemeat succeeded because it was supported by a number of other sources, all carefully choreographed by a central body. Agent reports, traffic analysis, newspaper reports and signals on channels known to be accessed by the Germans all fed supporting snippets, which together made up a convincing false picture of allied capabilities and intentions. With some judicious addition of erroneous reports just to make it more convincing. This was not cheap – although the main cost was in suppressing German access to any significant conflicting evidence. Note also that it was designed in the knowledge that it would not survive for long.
Contra Z’s point, intelligence is usually contested and reviewed – this made it easier for Cheney and Bush (just as one can always find an economist with the “right” views).
But JQ’s game theoretic claim is false – in a game of bluff where the parties have to coordinate internally and external visibility is hard to control, disinformation is neither cheap nor easy. As usual, the model is too simple to be useful.
Andrew F. 06.26.13 at 2:17 am
Tim, no one is claiming that raw information gathered is inherently reliable. It’s not – it can be unreliable even when no one has any motive or intent to be deceptive. As Hayden said at one point, “if it were a fact, it wouldn’t be intelligence.” Or as Mark Twain once put it, “the problem isn’t that people know so little, but that people know so much that just ain’t so.” That’s a fact of which practitioners and consumers of intelligence analysis are sharply aware.
But that doesn’t mean that we are incapable of sorting more reliable from less reliable sources with some degree of success, and using various means to assess the meaning of any particular bit of raw information. It certainly doesn’t imply that we’re so bad at deciphering raw information that the entire endeavor of intelligence is a wash.
The point of the examples brought up is not, as Consum rightly argued against, to establish by anecdote that intelligence analysis gets it right more often than not. Instead the examples are intended to show cases which have features that we intuitively recognize as being capable of generalization. Ultra and Magic are two examples of very broad signals intelligence programs that produced information of high reliability and value, and they have features to them that enabled the Allies to rely on them. Certain human sources, like Ames, or Penkovsky, also have features that enabled the information they passed along to be considered very reliable.
Snowden, actually, in some ways is a good example of a source of intelligence that also has features which may be generalized to other cases, which we recognize as justifying us in relying on the authenticity of certain documents he may pass along. To be clear, I intend nothing pejorative in this description of Snowden.
Now, there are all kinds of cases that are less clear cut, in which the possibility of disinformation contributes to what Angleton called the “wilderness of mirrors” effect (though Angleton obviously took the phrase from Eliot). And then there are other cases where there may be no disinformation, but the implication of the information gathered is unclear, particularly when it is being used to resolve “mysteries” (such as the intent and causes of a foreign government’s policy, or the future course of events) rather than “puzzles” (such as whether there are Soviet missiles in Cuba), to use Treverton’s terms.
I also agree that so called “open source intelligence” can be particularly valuable in answering some questions – sometimes more valuable than any other form.
I disagree that intelligence failures are inevitably covered up, Tim. In fact, as in most other human endeavors, imho the failures grab most of the postmortem analysis (along with most of the inquiries by Congress).
Anyway – my point is that the possibility of disinformation is a well known problem (for a horrifying example of how seriously it’s taken, read Heuer’s “Five Paths to Judgment”, on the treatment of Yuriy Nosenko and the internal struggle over his assessment), but not one that disables the entire endeavor of intelligence, any more than it disables us from sorting “intelligence” we gather from friends, family, neighbors, websites, periodicals, and broadcasts into various degrees of reliability.
jay 06.26.13 at 2:20 am
“But JQ’s game theoretic claim is false – in a game of bluff where the parties have to coordinate internally and external visibility is hard to control, disinformation is neither cheap nor easy. As usual, the model is too simple to be useful.”
Not really. For about a year the CIA/Special Ops/Air Force/sWetwork/Drone Boys were killing Jihadists and Taliban seconds after they got off the cell phone or sat phone. In some cases, while they were still on it. By the time of Tora Bora, the Talibs and Al Quida guys were using messengers. The busting of the email and online systems followed at the same pace, but with more arrests and fewer liquidations. Al Quida and the Taliban went signals dark pretty quick,
but for the next 8 years, colour coded “threat levels” ratcheted up day after day, week after week, based on vague and unspecified “intercepts”.
While some of these were probably the System Gaming the Media and Public for control of the message, some were probably disinformation ploys by Al Quida and the Taliban.
In Iraq, nary a Tikriti was caught by cell phone or sat phone. They used burners, and again, messengers, and the jihadi webs wern’t much use, because they U-tubed after actions, not before actions.
jay 06.26.13 at 2:37 am
Andrew F.
“I disagree that intelligence failures are inevitably covered up, Tim. In fact, as in most other human endeavors, imho the failures grab most of the postmortem analysis (along with most of the inquiries by Congress).”
Personally, I doubt that.
There has not been a US Crown Jewels expose since the Rockefeller Commission of 1975, and that got bogged down in the JFK/Zapruder film. Declassified documents released since then have show that the Commission was lied to, lead, misdirected and some programs were sacrificed to hide others.
Iran Contra was a joke. How about the Blow Back failures, Noreiga, Death Squads or Contragate. Basically, as the head of the NSA just proved again recently, you can lie to the Senate and Congress about anything with out consequence, as long as its not drugs and sports, or consensual sex.
Consumatopia 06.26.13 at 3:01 am
Good point. The cost of obscurity is especially great for democracies–we lose democratic accountability over our government. Since the purpose of security services is ostensibly to uphold our form of government, this is, in principle, self-defeating. It’s probably often the case that generating disinformation costs more than allowing yourself to be successfully spied upon.
jay 06.26.13 at 3:51 am
Consumatopia,
it depends on your status. If you are a major International power state, say, the US in the Bush II First Administration, then spreading disinformation and propaganda abroad, was bad for the State, because it fed back into the public decision making loops and lead to willful failures of public intelligence and decisions.
If you are a minor regional State, say Iraq, then it can keep local state vultures off your back, maybe long enough to reach a settlement with an International Broker or the International Power State targetting you.
John Quiggin 06.26.13 at 4:43 am
As you know, Stephen, I don’t believe that detailed factual knowledge is a major asset in reaching the correct conclusions on these issues. But you have consistently maintained the importance of factual detail, and as Ronan points out, you appear to have got the factual details wrong in this case. So, either an apology or else an explanation of why Ronan’s sources are wrong would be appreciated.
From my way of looking at these things, the commonality of the word “Special” is highly significant. In my experience, any occurrence of “special” in conjunction with police or security matters is cause for alarm.
jay 06.26.13 at 5:20 am
“From my way of looking at these things, the commonality of the word “Special†is highly significant. In my experience, any occurrence of “special†in conjunction with police or security matters is cause for alarm.”
In some ways, it is worse than that. While in media and common conversation the Special Branch is portrayed as an anti-terrorism, anti-spy, detective agency, each British Police Force has its own Special Branch, so in American parlance they are more like the NYPD Intelligence Division or the LAPD CTSOB.
Tim Wilkinson 06.26.13 at 2:13 pm
Peter T: in a game of bluff where the parties have to coordinate internally and external visibility is hard to control, disinformation is neither cheap nor easy
Far from being a refutation of JQ’s very specific game theoretic model, this would tend to support the general thesis that the Spy vs. Spy game is a massive negative-sum waste of resources, and that the best thing to do would be to de-escalate, instead of sustaining the massive self-serving clandestine operations industry by treating the whole business as literally a game which is to be accepted as standard practice.
The high cost of the whole (peacetime) secrecy and spying business, and of the general mistrust it both fosters and relies on is exactly the point. As JQ points out, one of the effects of eavesdropping is is to disrupt the communications of the ‘adversary’ by necessitating (though N.B. Andrew F’s claim that JQ considers this the main use of sigint in wartime is manifestly untrue and insupportable). This is the rationale Assange puts forward for encouraging leaks – to attack the internal communications network of huge conspiratorial organisations, thus forcing them to restrict communication in various ways and therby degrading their capability for large-scale co-ordination. Is this really the approach states should be taking toward each other?
The military history buffs, security industry flacks and gung-ho jingoists no doubt regard this hostile approach to other nations as a not-so-regrettably necessary component of a properly patriotic approach to international relations; for the rest of us these sour informers, these bate-bearing spies, are a insidous influence and a bar to greater co-operation, understanding and, yes, even peace.
If the idea is that misinformation and misdirection is not merely costly, but so costly that it’s unlikely to be engaged in so that a critical mass of pervasive uncertainty won’t in practice actually eventuate, I’d draw your attention to your own remark that the main cost was in suppressing German access to any significant conflicting evidence. Countermeasures of all kinds are costly and we’d be better off trying to move to a situation in which they weren’t relied on so much.
In any case, the extraordinary measures taken to try and make a particular piece of disinformation stick in the case of Operation Mincemeat is a different matter from routine obfuscation which aims to muddy the waters rather than actually convince one way or the other. The latter is what underlies JQ’s specific contention that as long as it is possible for counterspies to generate misleading information most of the time, spies are useless even when their information happens to be correct. If the defence plays optimally, the spymaster can never have any reason to believe one piece of information produced by spies and disbelieve another.
Peter Erwin 06.26.13 at 11:07 pm
(Extended comment, part 1 of 2):
I confess to being a bit confused as to what precisely John is objecting to (other than the excesses of secret police agences against their own people). Reading the 2001 article in which “the argument about the uselessness of spies is developed at much greater length” didn’t help, since once one strips away the literary name-checking, highly dubious history (spies were invented by novelists first, and then copied by governments?), and other irrelevancies (what, after all, did poison gas and tanks in WW1 have to do with spies?), there is very little of substance there on the “uselessness” of espionage. (Note that I’m not disagreeing with the arguments that out-of-control secret police are a very bad thing, or that hysteria about spies — or terrorists — can be exploited to enable the same and is thus also a bad thing.)
It’s true, for example, the Germans failed to set up any sort of network of reliable agents within the UK prior to or during WW2. Of course, they certainly tried, and their failure is probably a combination of German incompetence and British competence at counter-espionage. Which argues for the importance and utility of counter-espionage, at the very least.
But it hardly follows from this bit of cherry-picking that espionage, broadly considered, was useless to the Germans (let alone to the Allies). German codebreaking enabled them to read British diplomatic codes (including those used by the British at the 1938 Munich Conference; Hitler reportedly delayed one meeting with Chamberlain by an hour so that he could read the translations of Chamberlain’s most recent communications with London) and several of the British naval codes, which helped them successfully invade Norway, sink the aircraft carrier Courageous, and (most importantly) direct U-boat wolf packs to Allied convoys in the Atlantic — probably the closest the Germans came to threatening Britain with actual defeat. (There’s also the significant help Rommel received in North Africa in 1941 and early 1942 due to the fact that the Germans could read detailed and up-to-date reports on British forces and plans transmitted by the US military attache in Cairo — in part because the Italians and Hungarians had used classic “black-bag” espionage techniques to obtain US diplomatic codes.)
Of course, the Germans did indeed lose the war. (The fact that they had well-designed airplanes didn’t enable to win the war, either, which doesn’t exactly mean that airplanes are militarily useless.) The reality is probably that the Allies were simply much better, overall, than the Axis at intelligence work, and this is one of the reasons why the Allies won the war.
Peter Erwin 06.26.13 at 11:11 pm
Extended comment (part 2):
There’s some confusing concession to the idea of “codebreaking in wartime”, which John somewhat erroneously thinks “relies only marginally on traditional spying methods”. (There’s also the baffling claim that codebreaking somehow “disrupts enemy communications”, which as others have pointed out makes no sense at all, and isn’t explained or justified.) This is wrong in two ways, since (1) codebreaking has been a part of espionage since at least the Renaissance and is thus eminently “traditional”; and of course (2) actual codebreaking has historically benefitted from other “traditional” methods like paying foreign government employees to pass their secrets on to you, “black bag” breaking-and-entering operations, and so forth, as was the case for both the Allies (Enigma) and the Axis (the American military attache code).
This also seems to miss the important point that codebreaking “in wartime” almost always requires codebreaking in peacetime — all the major breakthroughs in WW2 codebreaking (the Germans reading British diplomatic and naval codes, the Polish/British success with Enigma, the American successes with the Japanese diplomatic and naval codes) came about because of efforts that were underway years before the war began. In the case of Enigma, this included very traditional espionage work in the form of the French recruitment of Hans-Thilo Schmidt, who David Kahn considers “WW2’s greatest spy.” (He puts Schmidt ahead of both Richard Sorge and Juan Pujol in terms of actual influence on the course of the war.)
Of course, carving out an exception for “nontraditional” things like codebreaking implies that there’s nothing wrong with the NSA and GCHQ reading everyone’s email and telephone conversations, since that sort of thing isn’t really “traditional spying methods”, either.
But then on the other hand we’re meant to admire and emulate Henry Stimson, who shut down the US State Department’s codebreaking operations in 1929. So codebreaking is simultaneously actually useful and something no one should do…. (Though we’re evidently not supposed to remember that the same Henry Stimson who became Secretary of War in 1940 did absolutely nothing to impede the US Army and Navy’s codebreaking operations.)
Peter T 06.27.13 at 2:03 am
re Tim @100
My point addresses JQ’s game theory. That disinformation is mostly neither cheap nor easy does not mean that it’s not worth it. The executive of a firm that hides various deficiencies from a prospective buyer may make a few tens of millions more, at comparatively trivial cost. The buyer that pays an informed insider to point to the deficiencies may save a few tens of millions, at comparatively trivial cost (and note that this is not necessarily a zero-sum game). Game theory does not tell you where the balance lies – only the details can do that. Sure we would be better off cooperating, but then we would not be arguing over the price, would we?
John Quiggin 06.28.13 at 4:14 am
@Peter E I agree that you are confused about my argument, but I’m unable to clarify it any further for you
@Peter T The game is indeed not zero sum. My point is that large classes of games involving espionage are negative sum, and negative in expectation for both players. One player may come out, but game theory can’t predict which – it can only say that with repeated plays, both will end up worse off than if they had co-operated
Comments on this entry are closed.