Benkler and Starr: What’s So Hard About Local? (Updated)
Posted by chanders on March 6, 2009
The gist for the Twitter generation:
- In the debate about the future of news production in a networked age, the question of the moderately local (or rather, the question of the neither “big” nor “hyperlocal”) remains the great unanswered question.
- The most promising candidates for peer-produced, commons based news production are long-term, national or regional, self-directed (enterprise), non-beat based reporting projects. Often, these come in the form of a special project with a high percentage of obviously modular, granular components (such as a detailed document review that could be outsourced to volunteers).
- Most local reporting, on the other hand, is deadline-driven, small-scale, reactive, and beat-based, and contains a questionable number of modular or granular components.
- The problem when it comes to replacing or supplementing local news production with commons-based labor thus has more to do with the organizational structures in which news work is embedded, rather than with the inherent nature of the work itself.
- There have been very few experiments with commons based peer production on a local level. Usually the organizations that have bothered to do the experimenting (like Philly Future, or Young Philly Politics, or Hallwatch) have been those organizations with the fewest resources. We need to stop relying on the starving visionaries. Or rather, somebody with cash and an organization infrastructure needs to take the lead and help the visionaries not starve.
- Paradox One: The bigger and more complex the newswork, the easier it is to share and distribute.
- Paradox Two: News projects most amenable to peered production are those forms of newswork by which reporters have most closely guarded their professional status. The most distributable forms of work may be those that take the form of Pulitzer winning articles, and 12-part investigative series. On the other hand, the smaller, less prestigious, under-recognized form of newswork may be the hardest to supplement through commons-based production.
- UPDATE: [I wanted to add a final point to this list. Could we make local news MORE distributable?] One way to make local journalism more amenable to a peer model would be to a) extend the time horizon of local newswork, b) expand the geographical basis of local news and c) make it less reactive and more enterprise. Can we make local newswork long term and de-local? Can we blow it out, in other words? What would this mean? Sadly, though, this seems to run in direct contradiction to the dominant mode of the web, which is favoring reactivity and speed> not the reverse.
Now, the long post:
In what we might usefully start calling the Benkler – Starr debate about the future of news production in a networked age (a conversation that was kicked off, I am happy to recall, at our own Columbia University’s Changing Dynamics of Public Controversies symposium), the question of the local remains the great bete noire. In his magisterial article in the New Republic, Starr notes that “the newspapers that seem most endangered by current trends are the ones in the middle–metros that do not draw substantial numbers of readers from beyond their regions.” Benkler, after chronicling the ways in which networked news production may begin to fill in some of the gaps left by dying newspapers, admits that “we do not have good research to know whether this [commons-based peer] system is also working for local politics and potential corruption as well.”
My own project began with the hunch that yes, it is the news economy “in the middle” that is the most in danger. I hope some of the work I’ve been doing might begin to address the research-gap pointed to by Benkler, above. In essence, what I want to do in this post is run my own findings about local newswork through the “Benkler algorithm” to try and answer these questions: what aspects local news production might be better carried out under a commons-based, peer production model? Why are there so few examples of these models on a local level?
First, some definitions. Benkler’s detailed descriptions of what he means by “commons-based peer production” can be primarily found in two places: the book, The Wealth of Networks, and in an article, “Coase’s Penguin, or Linux and the Nature of the Firm.” His arguments are complex, and I am going to start by posting an excerpt of his main point here (it should be noted, though, that if you understand what Jay Rosen is talking about when he speaks of pro-am journalism; what Jeff Jarvis is talking about when he talks about networked news, and especially projects like Off the Bus, than you already have a good idea what Benkler is saying:
Markets and firms use property and contract to secure access to bounded sets of agents, resources, and projects, thereby reducing the uncertainty as to the likely effects of decisions by agents outside the sets on the productivity of the agents within the set … Peer production relies on making an unbounded set of resources available to an unbounded set of agents, who can apply themselves towards an unbounded set of projects and outcomes …
What, then, are the limitations of peer production [for informational goods]?
… The incentives problem is a trivial one once a sufficient number of contributors can be involved. The actual limitation on peer production is the capacity of a project to pool a sufficient number of contributors to make the incentives problem trivial. This means that the modularity of an information product and the granularity of its components are what determine the capacity of an information product to be produced on a peer production model. If the components are sufficiently small grained, and the cost of connecting people to projects sufficiently low thanks to efficient cheap network communications, the incentives necessary to bring people to contribute are very small … This leaves the organization problem—how to get all these potential contributors to work on a project together, how to integrate their contributions and control their quality—as the central limiting factor.
[I also want to note here that this organization problem is exactly what Amanda Michel is talking about when she writes in CJR that "pro-am journalism demands a new kind of management."]
For Benkler, the key factors determining the limits of informational peer-production include: the size of the set of agents, the granularity and modularity of the labor in question, the size and nature of the set of resources and projects, and the problem of integration. Given all this, I wanted to look at some of the kinds of newswork I encountered during my Philadelphia research and see what qualities they possessed rendered them more or less amenable to a peer-based model of news production.
[I want to note, again, that Benkler never argues in his New Republic letter that all journalism can or should be peer produced; indeed, nearly all the examples in his letter combine commons-based and traditional organizational forms. I want to avoid the "curmudgeon tendency." Nevertheless, its a useful exercise to see where the model does and doesn't seem to apply.]
Examples of successful local networked reporting and organizing included the Philadelphia Independent Media Center’s coverage of the Republican National Convention protests, the Daily News helmed Next Mayor project [doc], the community website Philly Future, and the Norg’s conference. Apart from these cases, however, I saw few examples of networked news; indeed, some of the websites I just mentioned could barely bother to link out, never mind engage in peered reporting on a regular basis. What explains the difference?
First, but perhaps least importantly: the “local” nature of “local” news has the practical (if not theoretical) effect of reducing the total pool of possible agents who can contribute to a project. Duh. However, this is often overlooked — and it is also not the most important factor.
A second factor concerns the nature of the work in question. Is local newswork composed of enough finely grained modules to be distributed successfully in the context of peer-production? Sometimes. But even if we can break the practice of reporting the local news into enough modular and finely grained pieces …
Third, the type of newswork typically carried out by local reporters is constrained by the actions of institutions and systems that lie outside the control of those doing the work. Others agents, in other words, fix the boundaries of newswork, even if the set of journalists themselves are not bounded. Local newswork is often either reactive (insofar as it responds to an event that has already occurred) or monitorial (reporting on an already existing bureaucracy whose importance and behavior are predetermined).
Finally, there is the time horizon of newswork, or, in journalistic terms, the problem of deadlines. Even if there was an unbounded set of journalistic agents with access to an unbounded set of news resources and an unbounded choice of projects, the rhythms of daily journalism– rhythms which seem, if anything, to be increasing in speed– make the peer production of local news problematic at best.
In sum, the criteria limiting the peerability of local newswork has less to do with the nature of individual modules of work than it does with the nature of the larger projects in which those larger modules are integrated. Using Benkler’s criteria, the most appropriate venues for collaborative, peered reporting are:
long-term, national or regional, self-directed (enterprise), non-beat based reporting, usually in the form of a special project with a high percentage of obviously modular, granular components (such as a detailed document review that could be outsourced to volunteers).
These are all characteristics of newswork that stand in sharp contrast to a
deadline-driven, local, reactive, beat-based practices of local journalism that contain a questionable number of modular or granular components.
Examples of successful local networked reporting, such as Indymedia coverage of the Republican National Convention protests, or the Next Mayor project, or Philly Future are the exceptions that prove the rule; each example possesses enough of the necessary components called for by Benkler to render them amenable to less hierarchical and market-based organizational structures.
There’s an irony here: the news projects most amenable to peered production are those forms of newswork though which reporters have most closely guarded their professional status. The most distributable forms of work may be those that take the form of Pulitzer winning articles, and 12-part investigative series. On the other hand, the smaller, less prestigious, under-recognized form of newswork may be the hardest to supplement through commons-based production. The work of the career city hall reporter and the police bureau chief are most at risk in the digital age.
So what can we do?
First, and this is major — its important to keep in mind that there have been very few experiments with this stuff on a local level. Usually the organizations that have bothered to do the experimenting (like Philly Future, or Young Philly Politics, or Hallwatch) have been those organizations with the fewest resources. We need to stop relying on the starving visionaries. Or rather, somebody with cash and an organization infrastructure needs to take the lead and help the visionaries not starve.
Second, it seems clear that the problem with local peer production of news has less to do with the the granularity or modularity of the work itself (though there do seem to be some forms of local newswork that are NOT very granular, like the beat system) and more with the forms and processes in which they are embedded. One way to make local journalism more amenable to a peer model would be to a) extend the time horizon of local newswork, b) expand the geographical basis of local news and c) make it less reactive and more enterprise. Can we make local newswork long term and de-local? Can we blow it out, in other words? What would this mean? Sadly, though, this seems to run in direct contradiction to the dominant mode of the web, which is favoring reactivity and speed, not the reverse.
Third, if research shows that there are forms of newswork that a) are suffering from market failure and b) are harder to network and peer, than it stands to reason that the public policy interventions must be directed here. Once again, this goes against the dominant trends. What are the easiest projects to fund on a non-market basis? Big projects, long-term projects, complex projects. But these may be the kinds of news that need the least help.
Where should our foundation dollars go? Perhaps they should go towards assuring that the so-called “lowly” (and yet, so oddly difficult to fund either a peer-produced or market based substitute for!! so much for lowly!) beat reporter, police bureau chief, crime reporter, city hall reporter, can survive.