Document-Sourced Journalism as a Rent-Seeking Industry

NOTE: This unfinished post has been sitting in my draft box for about two months. With each passing day there are diminished motivations to complete it. So I am just going to hit the publish button.

Three years ago I wrote this little piece that touched on the extant difficulties of successfully replicating a distributed wikileaks model of document-sourced journalism. The crux of the problem: social media by and large fails as an editorial platform for document-sourced journalism. To be more specific, editorial capacity does not spontaneously emerge from social media as a solution to a type of collective action problem for wide-scale information/knowledge dissemination of document dumps. Wikileaks, simply as a technological platform for anonymous distribution of leaked documents, does not produce information/knowledge shocks. To achieve the latter requires “wikileaks” to also function along the lines of a traditional journalistic organization. That is, it must provide editorial context, expertise and discretion1.

But this transition of Wikileaks from technology platform to journalistic organization is what turns it into a “firm.”2 Now what I mean by this is further elucidated in the footnote below. But “Wikileaks as a firm” introduces an attendant issue of “geopolitical differentiation.” And it is the problem of geopolitical differentiation that prompted me to originally comment on the non-triviality of replicating a wikileaks model.

I bring this all up because the recent leaked announcement of a Glenn Greenwald,Jeremy Scahill Laura Poitras journalistic media venture financed by Pierre Omidyar looks to me to be the first actual legit successor to wikileaks. However, this announcement has not been greeted by universal acclaim in the radical community. Perhaps the general nub of this dissent is captured by this Arthur Silber rant. Interestingly, I find the radical critique to mirror the statist one. The crux of each is a challenge to the editorial discretion of the journalist in question. The statist claims too much is being leaked(usually attended by gobbledygook about how self-appointed editorial discretion is circumscribing the legitimate actions of democratic collective action) while the radical denounces too little has been divulged(accompanied by accusations that self-appointed editorial discretion confers a type of power that turns the dragon slayer into the dragon).

But there is a straight-forward rejoinder to the radical bemoaning the lack of an objective standard regarding what is published and the pace at which it is published. Namely, the leaker, Snowden(and before him, Manning), did not leak to a technology platform or to a social network, but rather to an entity the leaker self-calculated would provide maximum information shock value(“news”). Any radical ire on this matter should be directed at the leaker because it was the leaker who made the risk-reward calculation that the risk was only worth it if the reward entailed wide-scale publicity of the information being leaked. Information shock value requires editorial discretion on the part of the journalist(or journalistic entity). And as elaborated in the appending footnote, this editorial discretion entails a type of firm organization pattern. Sans this information shock value–which is what the Firm is meant to produce–the Firm offers little incentive to the leaker to leak(the leaker is incentivized by information shock) and without the participation of the leaker, the Document-sourced journalistic firm is a null entity.

Frankly, the radical disputes on this matter illustrate an object lesson why you have firms. A generic/informal social graph3 not oriented around a specific objective with no dispute resolution implementation in place suffers from the bargaining costs of haggling(or, in this case, bickering) horizontal trading partners(connected nodes). Now this is not to say that all social organization stratifies around a firm or that a generic social graph cannot reproduce the same results as a firm, but in a certain class of cases, a firm will beat an informal graph and beat it badly. Document-sourced journalism is likely one of those cases. A leaker, working against a significant risk calculation, is likely not going to leak to a platform or a network that poses uncertainty with regards to editorial discretion. An informal social graph characterized by heterogeneous opinion regarding editorial discretion is not reliable(particularly from the point of view of the leaker).

1The original wilikleaks model treated the dissemination of information shocks(‘news’) from its document repository as a type of social media collective action problem. However, social media as a network connected graph(more often than not) organizes around communicating moral judgements than coordinating expert division of labor (In other words, better at opinion journalism than investigative journalism). Think of a “blog roll” or “twitter friends.” A typical blog roll does not consist of an assortment of cryptographic experts, language experts, cultural experts, policy and government department experts, etc… the type of graph relationships that could facilitate spontaneous editorial contextualizing of an arbitrary document D if node i in the given social graph decided to go document browsing.

Instead, each node in a social media graph, with regards to document browsing, roughly encounters the same “rational ignorance” problem inherit in the “voting problem.” Now if you know beforehand that it is “worthwhile” to spend resources examining a given document D, because it exposes important information, then you would do it. But if you don’t know beforehand, you probably won’t. The resources you expend to research if the document is indeed newsworthy, or the resources required on your part to contextualize a given document’s particular newsworthiness(which likely requires cross-discipline expertise), exceeds the benefit you can expect from your own considerable effort. Your “social graph” isn’t helping you out any more than your traditional circle of friends helps you out with respect to the traditional voting problem.

Now, if editorial context provides advance knowledge, then the problem outlined above perhaps goes away. In much the same way that the “voting problem” can go away in a given instance if a newspaper editorial exposes that in a race between candidate A & B, candidate A, say, turns out to be a Nazi sympathizer(rational ignorance potentially is mitigated by the strength of the information shock). If editorial context for a document is provided, then a social graph or network can form for the dissemination of the document as an information shock.

The above highlights the distinctions between a social graph and a “firm.” A Firm can certainly be cast as a type of connected graph, so it is a special type of social graph. However, a firm’s “connectedness” is oriented around the execution of a specific objective(or set of objectives), and a firm implements a dispute resolution pattern for the connected nodes to facilitate execution of said objective. Hence, a Firm can be thought of as a DRO subclass of a social graph.

Wikileaks, in its second iteration, transitioned from a technology platform to a firm. That is, the aforementioned collective action failures of social media to provide editorial context for leaked documents would be remedied by Wikileaks itself acting as editor. The role of an editor is to provide context and expertise for published news and to exercise discretion in regards to what is published and the time-frame that it is published in. This introduces all the classic hallmarks of a DRO type.



2 Note: A Firm does not necessarily mean “corporation.” A corporate legal entity is a subclass of a Firm



3Note: A generic/informal social graph does not necessarily mean peer-to-peer(P2P). Fundamentally, P2P is a type of protocol, in particular, a “discovery protocol” that offers an alternative to the more traditional client-server model. However, for some reason, quite a bit of “moral evangelizing” is attached to it because supposedly it is more “hierarchically flat.” But this is a bit of a misnomer.

Example: Bitcoin Transactional Verification(“mining”) operates via a P2P discovery protocol, but bitcoin mining nonetheless is dominated by firms. A typical “blog ring” operates over the standard client-server model but the social graph of the blog-ring itself may be hierarchically flat.

A P2P discovery protocol operates over a scale-free network exhibiting a power law topology(“the internet”). Power laws in internet topology are about as scientifically reliable as observing gravity to be an attractive force. “Six degree of separation” and the “80-20 rule” are cornerstones of what we might term “the small network.” The small network is neither horizontally flat nor decentralized. The “small” refers to the distance or hops between any two arbitrary nodes.

A good example of a hierarchically flat, decentralized document-exchange network is the library system from, say, two centuries ago. Flat, decentralized networks are generally non-interoperable with one another by default. So a document exchange between a node i in network A and a node j in network B likely entails significant transaction costs. Hence, cross-network document exchange is costly and limited to the few. Not exactly egalitarian…
(Note: the origins of the portable document format originated from trying to address the problem of document exchange on disparate hardware/software platforms back in the late 1980s).

P2P aside, the actual point being made is that the abstraction of a social graph is generally not concerned with the underlying application protocol, although an internet topology characterizes the network where the social graph lives(for the most part). The terms “generic and informal” refers to the binding relationship between nodes.