The Wikileaks Data Dump

27 July 2010 at 10:56 am Peter G. Klein 9 comments

| Peter Klein |

I’ve been fascinated by the reaction to the Wikileaks release of 90,000+ classified documents related Afghanistan war. US and British (and Pakistani) authorities are predictably outraged, while critics of the war are encouraged that the disclosures could help turn the tide, as did the Pentagon Papers three decades prior. What interests me the most, however, is the massive size of the Wikileaks archive. As the Guardian’s Roy Greenslade remarked, this is “data journalism.” Wikileaks doesn’t analyze, synthesize, attempt to corroborate, seek alternative points of view, write up the inverted-pyramid lead, or do the other things respectable journalists are supposed to do; it just dumps the data and lets others sort it out.

Some find this approach distasteful. A Pakistani official said “these reports betray a lack of understanding of the complexities of the nations involved.” Well, sure. They’re raw data, nothing more. But isn’t sharing data, and not just analysis, a quintessential New Economy phenomenon? Don’t we have search and analysis tools, data-mining algorithms, page rankings, and other means to sift through the huge piles of stuff that constitute the long tail? Shouldn’t expert commentary and analysis be replicable? Many journals now mandate data-sharing. E.g.: “It is the policy of the American Economic Review to publish papers only if the data used in the analysis are clearly and precisely documented and are readily available to any researcher for purposes of replication. Authors of accepted papers that contain empirical work, simulations, or experimental work must provide to the Review, prior to publication, the data, programs, and other details of the computations sufficient to permit replication. These will be posted on the AER Web site.” Why should foreign-affairs reporting be different?

Entry filed under: - Klein -, Methods/Methodology/Theory of Science, Public Policy / Political Economy.

Overconfidence Foss and Klein Critique of Kirzner

9 Comments Add your own

1. David Gerard | 27 July 2010 at 1:13 pm

I’m not sure how interesting of a question this really is. I think the fact that the wikleaks guy can still see the light of day is probably more interesting than anything to me than anything that has been revealed so far. I liked this summation from Fred Kaplan:

Some of the conclusions to be drawn from these files: Afghan civilians are sometimes killed. Many Afghan officials and police chiefs are corrupt and incompetent. Certain portions of Pakistan’s military and intelligence service have nefarious ties to the Taliban.

If any of this startles you, then welcome to the world of reading newspapers. Today’s must be the first one you’ve read.
2. murray | 27 July 2010 at 3:35 pm

data vs opinion

what we usually get everywhere and legion are opinions and slanted views of the world – who can ever know truth

but pure raw data? wow!

it will be interesting just how well researched it will be – how many quality analyses are we likely to see?

there’s also this to note – as a computer science grad with almost 30 years of experience i know how to control the access to data – to the point where the only person in an organisation who would have the ability to dump so much data would be the head of the organisation – that lower level staff can do it for the US Armed Forces shows that they spend more time playing with high tech toys than managing their internal security

in a proper system, any leaked material would have unique data elements that would point directly to the source

unless of course the leak was by the head of the Afghan campaign

m
3. Peter Klein | 27 July 2010 at 8:44 pm

David, Kaplan seems to be missing the point. Unless he’s read all the documents, analyzed them for patterns, etc., how does he know what conclusions to draw?
4. srp | 27 July 2010 at 10:07 pm

Lots of competent observers have read vast swaths of these things and say that there is no new information for anyone who’s been following the war. Herschel Smith, at Captain’s Journal, for instance.
5. Thomas | 28 July 2010 at 3:38 am

I’d say this to Kaplan: Suppose Wikileaks leaked an exchange of emails between top Israeli and American officials about the IDF’s nuclear arsenal, which, let’s imagine, was enough to prove that (a) Israel has hundreds of nukes, (b) American officials are aware of this and approve of it.

Now, we might say “welcome to the world of reading newspapers” to such data since “everyone knows” that Israel has nuclear weapons. But such a leak would reconfigure (if not altogether smash) the “strategic ambiguity” that is today’s official policy. There may be similar information in the data on the Afghan war.

Some also argued that “Collateral Murder” didn’t tell us anything we didn’t already know. But surely it reconfigured the Pentagon’s rhetorical situation. Afterall, it had already released a report on the incident that was based on the (at the time unreleased) video. The most important leaks, I think, will not be those that give us new knowledge but those that shift the ground of the conversation.

We all “know” that officials lie about what is going on. A good leak, however, makes a particular lie impossible to maintain. It undermines a particular denial, invalidates a particular whitewash, etc.
6. Peter Klein | 28 July 2010 at 8:37 am

Thomas, I really like this comment, but can’t admit it, or Fabio Rojas will call me a pomo!
7. Thomas | 28 July 2010 at 4:04 pm

You’ve got to stop living in fear of Fabio all the time, Peter!
8. David Gerard | 28 July 2010 at 5:22 pm

Point taken, Peter, though in this case I would put money down that what we find will be closer to Burn After Reading than All the President’s Men

My enthusiasm for Kaplan was, perhaps, that I saw it after reading a “Top 7 Most Shocking Revelations” list. On the list, I saw one thing that hadn’t either read or presumed to be true. I certainly wouldn’t have qualified it as “shocking.”

Good point, Thomas. It is one thing to assume something to be true, and quite another to see it thoroughly documented. I guess the same things could have been said about the “Climategate” leaks, in that I pretty much assumed that’s what those guys were doing and I had seen some of that IPCC sausage being made, particularly with their treatment of uncertainty. It will probably reconfigure those researchers to send stuff on private email accounts.
9. Bergies | 10 December 2010 at 4:38 am

Kaplan is of course right in this case, since none of the documents leaked are classified as “topsecret” (the formal categorization). All the documents have been accessible by at least 1 million Americans – hence, there really isn’t much to make such a big fuzz about. Besides the funny descriptionsf state leaders, almost everything is to be found in newspapers. o