Printer Friendly

Post-industrial journalism: adapting to the present.

The Dilemma of Institutional Change

Over and over again the working journalists we interviewed, across a variety of publications and media types, lamented the inherent difficulty in shifting the directions of their legacy media organizations to meet the challenges of the digital age. Zach Seward, the former editor of outreach and social media for the Wall Street Journal and now a senior editor at the Atlantic Media business publication Quartz, told us that the very success of newspapers at doing what they do makes changing them difficult:
   The notion of adjusting course for an organization that is either
   still obligated to put out a daily print product or is otherwise
   very good and well oiled at a particular process feels as though
   the best an organization in that situation can do is make slight
   adjustments, if they are obligated to a production process that
   already exists. It's truly no small miracle that daily news
   organizations are able to produce what they do already, so 100
   percent of effort is expended on existing processes.


What we've called this "presence of process" doesn't manifest itself just when it comes to making big decisions. Indeed, the nature of institutional processes is that they are enacted on a daily, even hourly, basis. Process shapes what is and isn't possible, not just in conversations between reporters, editors and publishers, but in the very technological infrastructures that make the production of journalism possible. Tools put in place to manage process also put in place the assumptions used to design the tools.

Take newsrooms' content management systems (CMS). A CMS has a built-in idea of workflow--when and how content gets created, edited, checked and published. As a result, a CMS doesn't just help an organization manage its content in a particular way; it also deflects or even prevents it from managing it in ways that aren't built into the CMS.

The point is general, of course; all process exists to forestall alternatives, but CMSes are often at an extreme, because the requirements and assumptions are encoded in software and are difficult to argue with, or to override. As Anjali Mullany, a former online editor with the New York Daily News and now a social media editor with Fast Company, put it:
   The CMS and the project management systems are the crux of a lot of
   these [process] problems. Maybe 90 percent. Sometimes workflow and
   CMS aren't even compatible, or the CMS is inconsistent with the
   workflow. Or the workflow destroys the CMS. Look at any major
   organization, where it's multiplatform. It's not uncommon to see
   the same version [of a story] a few times. Or several reporters did
   the same story because they weren't communicating. The great,
   flexible CMS that will allow you to change your process over time
   does not exist. You should do this: try to find the one reporter in
   NYC who likes their CMS. This is a huge problem. If your CMS
   restricts you, it's going to restrict everything about the
   newsroom. The technology you're using is going to change what
   you produce.


The dilemma here is clear. We already noted that institutions can be defined as stable patterns and processes that allow collections of people and technology to accomplish more than they would as a mere aggregation of individuals. These institutional processes provide news organizations with many advantages vis-a-vis other political, social and corporate institutions they monitor. But these stable patterns, particularly when geared to particular production cycles that are themselves wrapped around particular technologies, can constrain news organizations as much as they empower them to report the news.

Matt Waite notes that the problem with large, hierarchical organizations is not that they discourage creative thought--a subtle and important distinction: "When working in a newsroom, [process is] a huge problem. But often in rigid hierarchies, working within constraints, we could have the greatest creativity. The problem was just getting someone to say 'yes.' Getting it to happen." He also noted that organizations with highly refined processes tend can make trying novel approaches politically difficult: "Newsrooms are still structured like the military. That makes it hard to do anything without stepping on someone's toes."

We can also catch a glimpse of the difficulty of institutional change by looking at how startup news organizations, though largely made up of veteran journalists and editors, navigate changes in process. Andrew Donohue recalls that when the Voice of San Diego began, "we were just doing what we did at newspapers, but online. Report through the day, wrap up at 7, then put it up on the site. We were not worried about constant updates."

We heard a similar story from a senior editor at the New York Times: "We were told effectively that the cuts meant doing more with less, one less person, no letup in the coverage. At no point were we ever asked by someone who had the technical capabilities or authority to actually change the tools or the ways we might use them: 'Let's look at what you have to do in a day and see how we can change processes.' This is what was so maddening."

At a smaller, nimbler organization like the Voice of San Diego, however, it was easier to shift this legacy process toward one that made a bit more sense in the current technological era. We had "a structured routine that slowly unwound as we got more people, and as social media came upon us. Now, our routine very different in that we both get our stories the traditional way, through sources and observing, but we have to decide how to present the story --a blog post, a daily, a three-month thing, a crowd-sourced form. So that's the biggest question these days."

The "process gap" is often most visible in work patterns tied to content management systems, because those systems exhibit a double conservatism. First, deploying a CMS represents such an enormous effort, the design of the technology typically reflects managerial choices about how employee workflow should work. Second, like Donohue's account of process at Voice of San Diego, CMSes are typically updated incrementally; products with print-centric daily rhythms that are adapted for the internet often feel like web-and mobile-centric features are an afterthought, because they often are an afterthought.

It is possible to get a sense of how misfit many existing production processes are by seeing what "digital native" CMSes and their attendant processes look like. To take one recent example, Vox, the publisher of several niche media sites, including SB Nation and the Verge, designed its own CMS from scratch. As Trei Brundrett, Vox's vice president of product and technology, put it in a public interview, "We map our development plan around the tools that our editorial and advertising teams tell us they need." This seems an obvious way to work, but it actually involves essential and rare skills: an editorial staff that can correctly characterize its needs; management that encourages editorial and technical collaboration; editorial and technical departments able to talk to one another; and a technical staff talented enough to create a working product that is simple and stable enough to be usable. The point here is not that every news organization should build its own CMS that's not possible and wasteful even if it was--but rather to illustrate how far print-centric tools are from fitting the new realities of news production.

The units of journalism are often tied to the logic of daily updates, a logic that does not always exist under conditions of digitization. In response to changing user expectations of time and timeliness, organizations need to rethink everything about how stories are organized and accumulate in the queue of news work. The newsroom assembly line is almost entirely anachronistic as a way of producing content to be produced for digital use, and it must be rethought.
Recommendation: Manage the Internet's Technological Demands

A failure to rethink workflow under condition of digitization can
often lead news organizations to suffering all the drawback of
digital processes while achieving none of the benefits. Some
commentators have referred to this worst-case scenario as the
"hamster wheel"--increasing demands on journalists' time and loss
of professional autonomy.

The hamster wheel is real, but many who discuss it mistake its
cause. We are not technological determinists who blame "the
internet" for the hamster wheel effect. Rather, we blame news
organizations themselves for adhering slavishly to old processes
under new technological conditions. In other words, technological
demands of the internet must be managed in order for the hamster
wheel to be avoided. Examples of how to manage the internet might
include a focus on intelligent linking rather than constant
aggregation and rewrites of already existing news, rotating "link
whoring" duty, as Gawker does, and many other process changes.

Recommendation: Be Able to Override Your CMS

Content management systems often embody ossified newsroom
processes. To the degree this is the case, the ability to subvert a
content management system can be a powerful strike against the
casual tyranny of impractical process. Journalists should prepare,
individually or in teams, to be able to override every step of
their CMS. With luck and persistence, these hacks and workarounds
can lay the groundwork for a more rational process in the future.

There is an analogy here with the design of medical information
systems. As hospital records have become digitized, there is
tension, as always, between security and access. A system that is
secure enough to prevent all misuse would end up preventing at
least some good but unpredictable uses as well. However, a system
that allowed all potential uses would do too little to secure its
contents.

The usual compromise is a "break the glass" function (analogous to
breaking the glass covering of an alarm bell). A doctor who
requests files that the system, for whatever reasons, says are not
accessible to her, can override the security, saying, in essence,
"My need for these files trumps the systems' security model." If
she does this, she then gets access to the files.

However, to do so, she must be logged in so the system knows her
identity, she must provide a rationale for why she is overriding
the system, and she is told that her override will be audited
within 24 hours. If her reasons for doing so are spurious, she will
be disciplined.

What we are recommending is the journalistic equivalent of "break
the glass" for overriding the assumptions a CMS makes about process
and control. If a journalist wants to bypass or override a
particular step, for reasons that seem justified and urgent, she
should be able to do so, provided she is sufficiently senior to
have internalized the local version of news judgment; that she is
identified to the system and willing to provide the rationale for
the override; and that she is willing to vouch for this rationale
when reviewed by management.

This opens the door to the possibility of errors of commission, of
course, errors that come from journalists doing something they
should not have done, but far too many CMSes force errors of
omission, which is to say errors that prevent journalists from
taking advantage of an obvious opportunity. By allowing journalists
to override their own processes as needed and with review, news
organizations can keep their desire for predictable workflow from
crushing the opportunity for novelty and initiative on the part of
their staff.

Recommendation: Embrace Transparency

As a counterpart to the power of hacking your process and working
around your CMS, news institutions should also make the new
processes they are using to generate quality journalism transparent
and systematizable by other organizations. In other words, when you
invent a process that works, you should "show your work" so the
same process can be used by other news outlets. ProPublica has been
an industry leader in this regard. While some news organizations
might fear that this kind of transparency will "aid the
competition," the fact remains that, for a century, news processes
were an open book. We see no reason that organizations cannot
continue to make money and get scoops in this new era, even when
they show their work.


Why Engage in Journalistic Work? Motivation and Institutional Impact

The fact that an increasing number of individuals contribute to the information ecosystem for free, or do so for reasons that do not strictly boil down to making money, has caused almost as much consternation in the media industry as has the question of pay walls. Early optimism about the ability of "citizen journalists" to transform the news business was quickly overtaken by both professional defensiveness and the economic crisis that enveloped the newspaper business (a crisis that had nothing to do with amateur production of content, but which was often lumped in with arguments about citizen reporting).

We will discuss the role amateurs and interested citizens play in the larger news ecosystem in the next section. For now, it is enough to argue that we think both sides of what is now a very sterile debate are missing the point. The role of everyday people in news production is an institutional question as much as it is an economic one. In general terms, the fact that at least some news producers contribute their labor for free means that a world of limited information has now become a world of overwhelming, often unprocessed, information. This poses a general challenge for news institutions: how to come up with new institutional processes and procedures to go from an information-scarce environment to one that is information-rich.

In more specific terms, one of the major dilemmas of amateur production is how to organize, rationalize, and systematize that production. It is not a coincidence that Amanda Michel, the former head of the Huffington Post Off the Bus project, began her career as an organizer rather than a journalist. As an organizer, Michel was well trained in understanding what amateurs and volunteers can do, what they can't do, and how to get them to work together for the benefit of a larger institution. How to manage amateur production can thus be tied to larger questions of how new entrants to the journalistic ecosystem might turn themselves from ad hoc networks to institutions. We now turn to that larger question.

Information and Impact (or, What Is Journalism For?)

Institutions provide certain key advantages when it comes to reporting news in the public interest: the kind of leverage, symbolic power, continuity and slack necessary to go toe-to-toe with other institutions: politicians, governmental agencies, businesses, schools, nonprofits, religious organizations. Yet the very same "systems of established and prevalent social rules" that help give institutions their heft also, in their inertia, serve to block necessary and needed change.

The solution to this paradox is not to abandon institutions. Nor is it to blindly stick with the institutions that have traditionally provided the best journalism in the past. Institutions are needed to do certain kinds of important things--but we need to reinvent the existing ones and to invent new ones. We need to focus on the way formerly ad hoc social arrangements become institutionalized, the barriers to such institutionalization, and the lessons and strategies for reporting the news that can be gleaned by watching this institutionalization take place.

There are two dilemmas of institutionalization at the heart of 21st-century journalism. The first, obvious and widely discussed since the 1990s, is the requirement for traditional news organizations to adapt to the internet, and the attendant difficulties they are having in doing so. The second, however, is less widely discussed: New forms of news production, from Andy Carvin's curated Twitter feeds to MapLight's database journalism to the stabilization of nonprofit web publishers like the Voice of San Diego or the Texas Tribune, have to become institutionalized, because without the virtues of institutions, albeit ones fitted to digital production, these new efforts will not be able to survive or to become persistent or powerful enough to discipline other institutional actors.

An example of a new, loosely structured digital journalism organization achieving some level of institutional stability can be found in the paradigmatic case of Talking Points Memo. We focus on TPM here, not because it has not dealt with its share of struggles and institutional challenges, but precisely because it has. Understanding the dynamic interplay between organizational challenge and institutional evolution is key to understanding the ways that the news media ecosystem is changing. Launched in 2000 by Ph.D. student and journalist Josh Marshall, the site was largely indistinguishable from the numerous "single-person" political blogs that were launched during the early days of the blogging revolution.

In 2002, the architecture of the site was fairly typical of the blogging genre at this early stage, with a "personalizing" photo of Marshall himself and a two-column setup (links in a narrow column on the left, major content in the middle of the page). Four years later, in 2006, the look and feel of the site illustrated the emergence of a very different organization. The picture of Marshall remained, but a far more structured page greeted readers.

Most importantly, by 2006 TPM was employing journalists, a process that began in 2005 when Marshall solicited money from readers to hire two full-time staffers; he raised $100,000 directly from the public. The right-hand column also linked to TPMMuckraker, an affiliated project that aims to do more original reporting and "muckraking."

By 2007, the architectural transition of Talking Points Memo was complete. The web page had come to resemble a full-time journalism operation, with boxes, links and different size fonts indicating different branches of the project and various editorial judgments about important news. The growth in staff continues apace; in 2010 it had 16 employees, and by 2012 it had 28. The site also received a significant financial investment in 2009 from the venture capital fund Andreessen Horowitz.

By looking at the arc of Talking Points Memo over time, we see the emergence of a non-institutional website in 2000, followed by an increasingly complex level of organizational structuring, staff growth and symbolic capital accumulation (the site won a Polk Award in 2008 for its coverage of the politically motivated firings of U.S. attorneys). While TPM is, by now, an "old" project in digital terms, it is useful for precisely that reason. Only by looking at the history of digital organizations on the web can we see how the story of journalism in the digital age is more than simply one of decline and birth. There is institutional stabilization as well.

Just as important, the story of Talking Points Memo represents the stabilization of a hybrid series of old and new journalistic practices, not simply the adoption of traditional reporting methods for the digital age. TPM was a pioneer in what is now known as iterative journalism, which it defines as the "using of tips, reporting, and explanatory writing from readers alongside original reporting to piece together wide-ranging stories." Although less is known about how TPM incorporates these practices into the 2012 iteration of its organizational structure, there is little doubt that the solidification of Talking Points Memo's institutional capacity represents the mainstreaming of a certain set of organizational practices.

A more proximate example unfolded over the summer of 2012, when Homicide Watch D.C. was threatened with shutdown. Homicide Watch, as described in Section 1, represents a fusion of traditional court reporting and novel technical infrastructure; it operates on a tiny budget; and the founders, Laura and Chris Amico, offer licenses for their platform to other news organizations. It is an ideal case for creating high value at low cost by rethinking process.

Nevertheless, by the summer of 2012, after two years of operation, Homicide Watch was threatened with shutdown, for two reasons. The first was that, despite the Amicos' offering the platform for licensing, few news organizations bit. Homicide Watch is so different from the "story-driven/should we report this?" model of the traditional crime desk that no existing organization could use the platform without altering its internal assumptions and processes as a side effect. The process gap made it far harder than the Amicos imagined to license their platform.

Despite this persistent difficulty, they kept the site going, running on a shoestring. Then came the second problem. Laura Amico, the reporter in the duo, got a Nieman Fellowship at Harvard. Faced with even the temporary departure of the founder, Homicide Watch had none of the advantages of large institutions--a deep bench of talent, employees with overlapping responsibilities who can pick up the slack, and so on.

It was only a last-minute Kickstarter campaign that enabled the hiring of D.C.-based staff for Amico to work with remotely that saved the site. This delays but does not solve the problem--small organizations like Homicide Watch are marvels of low-budget leverage, but they are also perennially threatened. To survive and spread their model, they will need to acquire more secure sources of funding, a larger and more varied staff, and more complex processes for managing that staff. They need, in other words, to become an institution.
Recommendation: Create "Startup Guides"

Starting a new news organization isn't as hard as stabilizing these
startups over the medium to long term. Because of this, successful
startups (such as Talking Points Memo, the Texas Tribune, West
Seattle Blog, Baristanet) should create publicly accessible
"startup guides" that can be used by emerging news organizations.

We also need to keep in mind that, because these organizations are
successful, their founders might have little time or interest in
devoting resources to explaining their success. They, after all,
have journalism to produce! For this reason, these organizations
and others like them should receive foundation money that will
allow them to engage in this "meta-reflection."


Understanding how new journalistic organizations stabilize themselves, and how in so doing they make a particular set of institutional behaviors seem like common sense, is a missing link in our attempts to understand the emerging news ecosystem. It is a financing gray area as well. Most foundation dollars are directed toward projects that can demonstrate a tangible "impact," which makes them less likely to help organizations engage in the boring, out-of-sight practices of institutional stabilization (things like setting up payroll systems, purchasing office space and providing employee health care, as well as training new employees and hardening institutional norms). Now that large national foundations like the Ford Foundation are increasingly investing in traditional media outlets such as the Washington Post and Los Angeles Times, investments in smaller, not-quite-new-but-not-yet-legacy outlets seem even less likely. The Washington Post received $500,000 from the Ford Foundation; it is not hard to imagine what Homicide Watch might be able to accomplish with a fraction of that money.
Recommendation: Rethink How to Deploy Funding

"Public" or noncommercial resources (including government and
foundation money) should be used primarily to helping organizations
institutionalize. Paradoxically, this is what these foundations and
the public sector appear the least comfortable doing, focused as
they are on demonstrating impact. Given the importance and
fragility of new players, there must be a rethinking of this
funding strategy in the foundation world.


When all is said and done, how are we to understand if news institutions whether old, new, or somewhere in between--are doing what they are supposed to do? How do we measure the success of these organizations? When success is primarily defined as "business success," the answer is simple although by that metric, the news industry has been in a tailspin for at least half a decade. Once we no longer define success as simply "making money" but rather as "making an impact on the world," however, our calculations change. There are many more ways of defining impact than there used to be, although the complexity of the question has correspondingly increased. To understand if institutions are working, we need to understand their purpose, and we need to measure the impact they are having on the institutions they monitor.

The question of "impact" has only recently begun to become a topic of conversation within news organizations and in the "future of news" conversation space. ProPublica has long been a leader in thinking about the actual impact of journalism, writing in its "about" page that "in the best traditions of American journalism in the public service, we seek to stimulate positive change. We uncover unsavory practices in order to stimulate reform." ProPublica adds that it does this "in an entirely non-partisan and non-ideological manner, adhering to the strictest standards of journalistic impartiality." It concludes by noting that "each story we publish is distributed in a manner designed to maximize its impact."

This would appear to be a noncontroversial mission. Surprisingly, however, it is one that is not publicly echoed by more traditional media organizations, although a desire for "impact" does undergird journalistic belief structures more generally. Often, news institutions will argue that they are there simply to "present the facts" and that questions of what those facts will do lie outside their purview. Journalistic institutions usually view the news consumer as an empty receptacle for public information who, when well-filled with the proper knowledge, will act in a variety of democratic ways. The impact of the news, in other words, comes not from the news producers but the news consumers, from the democratic citizens themselves.

It should be clear by now that this empty receptacle analogy for thinking about, to quote NYU journalism professor Jay Rosen, "what journalism is for" is obviously not an analogy to which we lend much credence. Instead, we believe that it is news institutions themselves that often do the most to advance positive democratic outcomes. Given this, it has become essential to understand exactly how news organizations make an impact, and for news companies to admit that they are in the impact business.

We are heartened by the announcement, in the summer of 2012, that the Knight-Mozilla Foundation will be placing one of its fellows at the New York Times specifically for the purpose of designing ways for news organizations to measure impact. "What we do not have are ways of measuring how a piece of journalism changes the way people think or act. We don't have a metric for impact," Aron Pilhofer, the newspaper's editor of interactive news, wrote on his blog.
   This is not a new problem. The metrics newsrooms have traditionally
   used tended to be fairly imprecise: Did a law change? Did the bad
   guy go to jail? Were dangers revealed? Were lives saved? Or least
   significant of all, did it win an award?

      But the math changes in the digital environment. We are awash in
   metrics, and we have the ability to engage with readers at scale in
   ways that would have been impossible (or impossibly expensive) in
   an analog world.

      The problem now is figuring out which data to pay attention to
   and which to ignore. It is about setting up frameworks for testing,
   analysis and interpretation that are both scalable and replicable.

      It's about finding that clear signal among the white noise that
   tells us whether our journalism is resonating or not, whether it is
   having the impact we believe it should. Helping us clear away the
   noise is the goal of our proposal to host a Knight-Mozilla fellow.


We hope that this step by the New York Times and the Knight-Mozilla Foundation will open the door to other news organizations thinking hard about what they do, and why it matters. Only if they start to think of themselves as organizations that "do things" in the world can we ever hope to understand the value of news institutions, and the ways we can replace the institutional value currently being lost in the digital tsunami of the early 21st century.
Recommendation: Assess and Value Impact

Make assessing impact, including job assignments and promotions,
part of organizational culture. Consider partnerships with
organizations that can provide information or insight into areas of
desired impact.


What New News Institutions Will Look Like

We've now discussed why institutions are essential to secure the proper functioning of a healthy journalistic ecosystem. We've also discussed an institutional paradox: The traits that make organizations successful during times of relative social stability can be the very traits that leave them unable to adapt to a rapidly changing organizational reality. Given all this, what would healthy news institutions look like in the 21st century? What kind of institutional arrangements should newsroom editors, corporate CEOs, rank-and-file journalists and future of news commentators demand?

We should note, right off the bat, that the news institutions of the future will be smaller than they are today; drawing from our earlier arguments, we acknowledge that staffing reductions, lowered budgets and a need to "do more with less" have become the "new normal" for journalistic organizations. We also think that news organizations will probably get new forms of funding from a number of sources, including some form of digital subscription, website advertising, social media-driven sales strategies (such as those adopted by BuzzFeed), foundation grants and governmental subsidies. It is not our intent to recommend any of these revenue sources over others, although we do note that certain forms of revenue generation make the institutional strategies we envision below easier, while other choices make the transition harder.

We want to argue that news institutions of the future, apart from simply being smaller and revenue agnostic, should have three defining characteristics. They will have a hackable workflow. They will embrace a form of what we call "networked institutionalism," and many of the largest, national journalism organizations should embrace local accountability journalism in partnership with local news outlets. Finally, news institutions will have to dramatically rethink what counts as "valid journalistic evidence," find new ways to evaluate this new evidence, and program these collection and evaluation process into their hackable workflows.

The Hackable Workflow

Currently, news production processes are designed around two imperatives. The first is that they rationally manage the generation, transmission, editing and production of content, and do so for as many simultaneous platforms as possible. The second imperative, related to the first and largely a legacy of the print/broadcast production process, is that this workflow management is designed to produce a single finished product that will be "consumed" once and then disposed of. Thinking about workflow this way (and, more importantly, managing the production and dissemination of content this way) makes sense only so long as this "create once/consume once" model holds.

Online, journalistic content can be produced, added to, altered and reused forever. To take advantage of this change, workflow will have to be altered to support these new technological and cultural affordances. Creating a workflow that reflects the more flexible production of digital content will have the secondary consequence of making rigid newsroom routines more "hackable."

The organizational breakthrough of the hacker-journalist lies not in being up to speed on the latest social media tools or even in being able to manage a thousand-column Google Fusion Table. Rather, the key insight of journalists raised on the rhythms of digital production and programming languages is the understanding that "content" is not used once and then discarded. Rather, content is endlessly reusable and should be designed for perpetual levels of iteration. In our interviews with working journalists, we were struck by the degree to which all news organizations remain trapped in a basic newsroom workflow that sees the ultimate goal of journalistic production as a singular, finished product. Rebuilt news institutions will design their workflow around a new, basic fact: News is never a finished product, and there is never a daily paper or evening newscast that sums up the work of the entire day.

This implies that news content, and the production of that content, will take iteration as its starting point. News products will have to be made as reusable as possible: on other platforms, on other devices, in new news stories, and even by other news organizations.

It also has another consequence: Newsroom CMSes will have to be designed to allow them to be broken. The obvious corollaries are that the act of choosing (or, in rare cases, designing) CMSes will have to include questions of who can override the expectations embedded in the CMS, and how, and that the processes put in place around the CMSes will have to emphasize the ability of at least some employees to exit the expected process in order to make novel decisions in response to novel circumstances.

In other words, they need to be flexible and adaptable to particular organizational needs. The focus of news production management should not be the creation of a final product within a one-size-fits-all workflow; rather, the focus should be upon the creation of endlessly iterable content through a highly hackable CMS.

The Networked Institution

Much ink has been spilled over the question of organizational partnerships in the news business, and many arguments have been advanced as to how institutions need to be more open to collaboration with other members of the digital news ecosystem. To date, however, the verdict on existing collaborative projects is mixed. A number of the New York Times' most highly touted collaborations (with the Chicago News Cooperative, the Bay Citizen, and the CUNY-sponsored Local, for instance) have come to a rather inglorious end; at the same time, many Times partner organizations have noted how working with a powerful organization has the potential to distort their own organizational priorities. The notion of institutional collaboration, while intellectually powerful, is in need of some rethinking.

We want to argue that the news organization of the future will probably not be an entirely open institution whose primary purpose is collaboration, nor should it focus on only collaboration in an entirely project-based sense. Instead, we'd recommend a strategy much like that pursued by ProPublica in its "Free the Files" project.

In Free the Files, ProPublica sought to crowdsource the collection of FCC political advertising buys. And because the media markets in question are inherently local, ProPublica essentially engaged in an act of local accountability journalism, even as it coordinated this journalism on a national scale. The final step for a project like Free the Files would be to collaborate with local news organizations to publish the data in relevant, journalistically interesting ways. This is neither permanent collaboration, nor is it based around a onetime event. Instead, it is using smart, targeted networked institutionalism to fill a gap opening up in local accountability reporting. Not surprisingly, this new collaboration is also based around the existence of new forms of journalistic evidence, specifically large data sets.

New Forms of Evidence

In Section 1, we discussed new skills that will be required of the postindustrial journalist. In many respects, these skills can be summarized as an ability to recognize, rather, evaluate and display new forms of journalistic evidence. What do social media conversations, large data sets, and on-the-scene, first-person media production all have in common? In essence, they present the 21st-century journalist with a plethora of new sources that can be integrated into the journalistic production process.

As we already argued, these changes in the larger media ecosystem present the individual journalist with new challenges and a need to master new skills. Every individual working in the news business thus needs to take this requirement seriously. At the same time, the institutions in which these journalists are embedded need to create organizations and newsroom workflow patterns that support individual journalists in this regard.

In other words, we cannot continually require reporters to master new skills and evaluative procedure without simultaneously providing them with a workflow and an organizational structure that shows them that such skill mastery is valued and rewarded. Such a workflow will need to be both hackable and networked in smart, labor-enhancing ways.

Conclusion: Journalism, Institutions and Democracy

In a 1995 essay, the late communications theorist James Carey writes eloquently about what he calls the Fourth Estate view of journalism, a view of the relationship between the media and democracy that did not emerge fully until the 1960s and the Watergate era:
   In this view, journalists would serve as agents of the public in
   checking an inherently abusive government. To empower it to fulfill
   such a role, the press had to possess special rights to gather
   news. Thus, under a fourth estate model a free press essentially
   was equated with a powerful press possessing special privileges of
   newsgathering.


Under the Fourth Estate view, according to Carey, the press increasingly began to see itself as the public's representative within the political arena. For this notion of representation to resonate, however, the public not only had to see the press as its authentic political stand-in, but also had to believe that this representative press was capable of accurately understanding and portraying the basic empirical reality of the world. It is fair to say, if surveys of trust in journalism have any validity at all, neither of these conditions holds true in 2012.

What Carey did not consider--what almost no one considered in the world of 20 to 30 percent newspaper profit margins that still prevailed less than a decade ago--was that the press might also become incapable of fulfilling its end of the newsgathering bargain. From the 1960s on, most media criticism consisted of the argument that journalism was capable of far more powerful, in-depth, aggressive newsgathering than it decided to undertake. As Downie and Schudson argue in their analysis of "accountability journalism," and as the 2011 FCC report on community information ecosystems reiterates, the problem with news today is as much one of incapacity as it is of purposeful neglect. We have also analyzed the connection among institutional capacity, the problem of time, and beat reporting in our discussion of David Simon's arguments: In short, much of journalism's value added lies in the operation of the daily routines, this monitorial beat system is best facilitated by healthy institutions, and institutional decline is leading to the evisceration of the unique journalistic resource.

At this point, a brief discussion of the economics of the news business is unavoidable, because it is at this moment that the future of news consensus breaks down. According to at least two camps in this debate, better market mechanisms will lead to revived institutional health, although the manners by which these camps define "better market" are directly opposed. A third perspective despairs that a market-based solution to the problem of institutional news industry decline can be found.

The first cluster of thought, represented by future of news thinkers like Jeff Jarvis, believes that the digital news ecosystem itself represents a more transparent, accurate marketplace than the monopoly news market of the previous regime. The contention here is that the funding for public interest journalism will emerge from a combination of transparency, increased public sharing and improvements in the ability of the advertising industry to micro-target consumers. Pointing to the monopoly status enjoyed by the most powerful news institutions for nearly a quarter-century, these thinkers see the current moment of informational abundance, the ability to tailor content to consumers, and "frictionless sharing" as remarkable steps forward from an earlier, less free model of media production.

David Simon, in comments on the blog post discussed above, nicely articulates a second understanding of what a "better" market means, one apparently shared by an increasing number of news industry executives. "I believe that local news can be sustained through an online revenue stream," Simon argues. "But it requires that institutional journalism value and protect its own copyright and act as an industry to protect that copyright. And further, it requires a real reinvestment in that product." To this list Simon adds the imposition of pay walls, which have, he contends, already demonstrated their success at the New York Times. In short, Simon and those like him argue that unified action to crack down on aggregators and charge for news will address the causes of newsroom decline via industry-based solutions. To maintain news organizations' position as the dominant provider of news, speed bumps should be installed on the internet.

A third perspective despairs that either of these market-based solutions can be easily conjured up. Thinkers and writers in this camp point out just how unusual the confluence between wealthy capitalist institutions and the public-minded journalism they produced actually was. They argue that digital market dynamics actually punish institutional players that seek to create broad-based, monitorial media content. Unlike thinkers in the second camp, however, they do not believe that the current dynamics of the digital news system can be easily undone, nor do they think the dynamics necessarily should be undone even if such an option were possible. Some thinkers within this perspective move from here to an argument that the public goods produced by news institutions (particularly beat reporting) can be funded only via non-market forms of subsidy, whether philanthropic or proceeding more directly from the state.

The three authors of this paper would place themselves in this third category, a standpoint that also informs our transition from institutions, in this section, to the news ecosystem that immediately follows in Section 3. We must move away, in other words, from pinning our democratic hopes entirely on the Fourth Estate conception of the press. Public accountability must come, in part, from the networked news ecosystem itself. Let us be clear: This is not to argue that these news networks exist in some sort of institution-free vacuum. Indeed, journalism institutions turn out to be some of the most important nodes within this new digital environment. Nevertheless, they must coexist in new ways, alongside and in concert with more groups and institutions than ever before--not simply for economic reasons but also for democratic ones. They must lean on these new groups and networks in new ways. We are echoing here our opening argument that the journalism industry is dead but that journalism exists in many places.

In the essay quoted earlier, James Carey contends that the "watchdog notion of the press, a press independent of all institutions, a press that represents the public, a press that unmasks interest and privilege, a press that shines the hot glare of publicity on all corners of the republic, a press that searches out expert knowledge among the welter of opinion, a press that seeks to inform the private citizen, these are ideals and roles that have served us well through some dark times." But, he continues, "as the century progresses, the weaknesses of modern journalism have become increasingly apparent and debilitating."

Carey's thoughts on the benefits and weakness of the Fourth Estate are as true now as they have ever been. The crisis, however, is even more acute than it was when he wrote those words in 1995. The communicative universe, moreover, has changed radically. If the democratic accountability fostered by the institutional press is to survive in a post-Fourth Estate world, democratic accountability must itself become a networked property.

Section 3

Ecosystem

The only reason to talk about something as abstract as a news ecosystem is as a way of understanding what's changed. The most significant recent change, of course, is the spread of the internet, connecting our computers and phones in a grid that is global, social, ubiquitous and cheap. As new capabilities go, the ability for any connected citizen to make, copy, alter, share and discuss digital content is a lulu, upending many existing assumptions about news and about media in general.

The news business in the 20th century was a fairly linear process, where reporters and editors would gather facts and observations and turn them into stories, which were then committed to ink on paper or waves in the air, and finally consumed, at the far end of those various modes of transport, by the audience.

A pipeline is the simplest metaphor for that process, whether distribution of news was organized around the printing press or the broadcast tower. Part of the conceptual simplicity of traditional media came from the clarity provided by the near-total division of roles between professionals and amateurs. Reporters and editors (and producers and engineers) worked "upstream," which is to say, as the source of the news. They created and refined the product, decided when it was ready for consumption, and sent it out when it was.

Meanwhile, the audience was "downstream." We were the recipients of this product, seeing it only in its final, packaged form. We could consume it, of course (our principal job), and we could talk about it around the dinner table or the water cooler, but little more. News was something we got, not something we used. If we wanted to put our own observations out in public, we needed permission from the pros, who had to be persuaded to print our letters to the editor, or to give us a few moments of airtime on a call-in show.

That pipeline model is still central to the self-conception of many institutions in the news business, but the gap between that model and the real world has grown large and is growing larger, because the formerly separate worlds of the professionals and the amateurs are intersecting more dramatically, and more unpredictably, by the day.

The main effect of digital media is that there is no main effect. The changes wrought by the internet and mobile phones, and the applications built on top of them, are so various and pervasive as to defeat any attempt to understand the current transition as a single force or factor. To understand this as a change to the ecosystem, it helps to have a sense of where the changes are showing up, and how they interact.

Here are a few surprises in our little corner of the 21st century:

* In 2002, after Senate Minority Leader Trent Lott praised Strom Thurmond's segregationist 1948 campaign, one of the people who did Lott in was Ed Sebesta, a historian who had been tracking racist statements made by American politicians to segregationist groups. Shortly after Lott said his praise had been an uncharacteristic slip, Sebesta contacted Josh Marshall, who ran the blog Talking Points Memo, to share similar (and similarly racist) comments made by Lott dating back to the 1980s.

These comments undermined Lott's ability to characterize his comments as a slip and led to his losing his Republican leadership position. Sebesta had built the database of racist speech on his own, without institutional support; Marshall was an amateur blogger (not yet having incorporated); and the source contacted the news outlet, 1,500 miles away, rather than vice versa. Indeed, as mentioned in Section 2, Talking Points Memo became the institution it is today because of what Marshall was able to do as an amateur (another example of institutional stabilization).

* In 2005, the London transit system was bombed. Sir Ian Blair, the head of London's Metropolitan police, went on radio and TV to announce that the cause had been an electrical failure in the underground. Within minutes of Blair's statements, citizens began posting and analyzing pictures of a bombed double-decker bus in Tavistock Square, and in less than two hours, hundreds of blog posts were analyzing this evidence. These posts reached hundreds of thousands of readers and explicitly contradicted Blair's characterization.

Seeing this, and overriding the advice of his own communications staff, Blair went on air again less than two hours later to say that it had indeed been a bombing, that the police didn't have all the answers yet, and that he would continue reporting as they knew more. When he spoke to the public, Blair had the power of all the traditional media behind him, but it was clear that merely having a consistent message on every broadcast channel in existence was no longer the same as having control.

* Starting in 2010, in a series of reports called Dollars for Docs, ProPublica covered the flow of payments between the pharmaceutical industry and prescribing physicians. It was a story that had been covered before in bits and pieces, but ProPublica brought several things to its investigation not previously seen, including a database it assembled from data the pharmaceuticals were required to make public, along with the ability and journalistic will to mine that database.

Dollars for Docs was not just a new report. It was a new kind of reporting. Though much of the data used were publicly available, they had not been centralized or standardized in a form that could make them useful; armed with this database, ProPublica has been able to report on a national story, while also providing tools for other organizations to cover the same issue as a local story; as of this writing, it helped spark stories in 125 other publications. (As a nonprofit, ProPublica can be both a news retailer and wholesaler.) In addition, it has been able to make its database as local as any news story can ever get: individual users can type the name of their doctor into the database and get a customized report. The harvesting and organizing of publicly available data thus became a platform for national, local and personal reporting.

Better access to individuals, as with Ed Sebesta; crowds, as with the London bloggers; and machines, as in Dollars for Docs, are driving working models that would have been both unthinkable and unworkable even 10 years ago: Huffington Post's Off the Bus project, covering every Iowa caucus in 2008 with citizen journalists, would have bankrupted the organization had it been done with stringers. The Guardian decided to crowdsource the tracking of expenses by UK members of Parliament, because the job, done by employees, would not just have cost too much but taken too long.

Journalists have always used tip lines and man-in-the-street interviews, and members of the audience have always clipped and forwarded favorite articles. What's new here isn't the possibility of occasional citizen involvement. What's new is the speed and scale and leverage of that involvement, the possibility of persistent, dramatic amounts of participation by people previously relegated to largely invisible consumption. What's new is that making public statements no longer requires pre-existing outlets or professional publishers.

Tip lines worked well only in geographically local areas, but NY Velocity was able to reach halfway around the world to get its critical interview in the Lance Armstrong doping case. Man-in-the-street interviews are random, because the professionals controlled the mode and tempo of public utterances, but with Flickr and weblogs, British bloggers could discuss the London bombings in public, at will, and with no professionals anywhere in sight. Dollars for Docs took disparate data and turned it into a database, which gave ProPublica an ongoing resource that was reused by it, other organizations, and millions of users over the course of two years and counting.

This is a change in degree so large, in other words, that it amounts to a change in kind. As Steven Levy observed, writing about the iPod, when you make something 10 percent better, you've made an improvement, but when you make something 10 times better, you've created a new thing. New digital tools can accelerate existing patterns of news gathering, shaping and publishing so dramatically that they become new things.

We are living through a shock of inclusion, where the former audience is becoming increasingly intertwined with all aspects of news, as sources who can go public on their own, as groups that can both create and comb through data in ways the professionals can't, as disseminators and syndicators and users of the news.

This shock of inclusion is coming from the outside in, driven not by the professionals formerly in charge, but by the former audience. It is also being driven by new news entrepreneurs, the men and women who want to build new kinds of sites and services that assume, rather than ignore, the free time and talents of the public.

The importance of news isn't going away. The importance of dedicated professionals isn't going away. What's going away are the linearity of the process and the passivity of the audience. What's going away is a world where the news was made only by professionals, and consumed only by amateurs who couldn't do much to produce news on their own, or distribute it, or act on it en bloc.

This a change so varied and robust that we need to consider retiring the word "consumer" altogether and treat consumption as simply one behavior of many that citizens can now engage in. The kinds of changes that are coming will dwarf those we've already seen, as citizen involvement stops being a set of special cases and becomes a core to our conception of how the news ecosystem can and should function.

Ecosystems and Control

To talk about a "news ecosystem" is to recognize that no news organization is now, or has ever been, absolute master of its own destiny. Relationships elsewhere in the ecosystem set the context for any given organization; changes in the ecosystem alter that context.

This paper began with a focus on the individual journalist, and on the various ways she can gather, process and make sense of information and events vital to public life. Most journalists do their work inside institutions; those institutions are shaped by everything from the size and makeup of the staff they employ to their self-conception and source of revenue. These institutions in turn shape the work of the journalist: which stories she can and can't pursue, what is considered good or bad work, who her collaborators can be, and what resources are at her disposal.

Those institutions are themselves in an analogous position, operating in the media environment that covers the news (and sometimes even the part that doesn't). This news ecosystem (hereafter just "ecosystem") is made up of other institutions--competitors, collaborators, vendors and suppliers--but it is also made up of the ways other actors affect those institutions. The audience's preference for news about Hollywood over Washington, the presence of the competition just a click away, the Supreme Court's current interpretation of the First Amendment, and the proliferation of high-quality cameras on mobile phones are all part of the news ecosystem of the early 21st century, the effects of the ancient and modern all mixed together.

The ecosystem also shapes institutional capability: the kinds of stories that do and don't get pursued are affected by everything from audience and advertiser desires to narrative frames. Everyone knows how to tell the story of a cheating athlete or a business gone bankrupt, but there is no obvious narrative frame for the tension between monetary and fiscal union in the EU, even though the latter story is by far the more important. Similarly, the facts and assumptions around things like access to data, validity of sources, the nature and limits of acceptable partnerships, and so on affect what institutions believe they can and can't do, and should and shouldn't do.

In the pipeline model of news, the existing institutions could be thought of as a series of production bottlenecks, owned and operated by media firms, and from which they captured income from both advertisers and audience. These bottlenecks were a byproduct of the incredible cost and difficulty of reproducing and distributing information, whether via printing press or broadcast tower. As noted in the last section, this was an ecosystem in which the institutions themselves had a high degree of control over their own fates.

A large, competent staff was required to print and deliver a daily paper; an even larger one was required to make and broadcast a news program. These costs and difficulties limited competition, as did the geographic range of delivery trucks and broadcast signals. Within the small numbers of organizations that could create and distribute news, whole professional structures arose.

Newspapers and magazines saw this institutionalization first, of course; the printing press preceded not just radio and movies but also steam engines and telegraphs. The entire professional edifice of writers and editors and publishers and, later, illustrators and layout artists and fact checkers and all the rest of the apparatus that went into creating a newspaper were built around and often quite literally on top of--the giant machines that put the ink on the paper. Radio and TV news departments followed the same pattern, inventing professional categories and practices to subdivide and systematize both the work and the categories of employment that went into making broadcast news.

Then came the internet, whose basic logic--digital replication, universally available, with no division of participants into producers and consumers--is at odds with the organizing principles of news production as it has existed since the 1600s. Abundance creates more disruption than scarcity; when everyone suddenly got a lot more freedom, every relationship in the old "charge for operating the bottleneck" model was up for grabs.

The arrival of the internet did not herald a new entrant in the news ecosystem. It heralded a new ecosystem, full stop. Advertisers could reach consumers directly, without paying a toll, and it turned out many consumers preferred it that way. Amateurs could be reporters, in the most literal sense of the word--stories from the Szechuan quake to Sullenberger's Hudson River landing to Syrian massacres were broken by firsthand accounts. The doctrine of "fair use," previously an escape valve for orderly reuse of small amounts of content among a small group of publishers, suddenly became the sort of opportunity that whole new businesses of aggregation and re-blogging could be built on top of. And so on.

When changes are small or localized and existing institutions are well adapted to those conditions, it doesn't make much sense to think about things as an "ecosystem"--simply responding to competitive pressures and adapting to small and obvious changes is enough. For institutions that produce news, however, the changes of the past decade have not been small or localized.

A common theme in writing about the response to those changes by traditional news outlets is the failure of newspaper management to recognize the problems they would face. This, in our view, misdiagnoses the problem: The transition to digital production and distribution of information has so dramatically altered the relations among publishers and citizens that "stay the course" has never been an option, and, for the majority of the press that was ad-supported, there was never an option that didn't involve painful restructuring.

A similar theme has been unpredictability and surprise, explaining the current crisis with the rationale that recent changes were so unforeseeable and have transpired so rapidly that traditional organizations were unable to adapt. This view is also wrong: There were coherent predictions of the trouble the internet would cause for the news industry going back to the late 1980s, and despite frequent invocations of "internet time," the pace of this change has been glacial; dated from 1994 (the first year of the broadly commercial web), management has had 75 consecutive quarters to adapt.

Individual accounts of even successful adaptation to the current ecosystem make it clear how hard such adaptation is. To take one example, in August 2011, the New York Daily News launched innovative live coverage of Hurricane Irene, replacing the front page of its website with a live blog, Storm Tracker.

The News then dispatched reporters out into the city, armed with cameras and phones (often the same device) to document everything from the evacuation efforts, to residents' struggles to shelter in place, to the effects of the wind and water itself. These live reports were interspersed with messages from weather services, emergency services and city government, all unfolding along with the storm.

The News' effort in live disaster blogging was a triumph, for which the News rightly won considerable praise. Also, it almost didn't happen. The precipitating event for Storm Tracker was not a new web strategy but the failure of an old one. The News building is on Water Street, in a Class A flood plain, so the police severely limited the number of workers who could go there on the weekend Irene blew in. This would seem to be no problem for filing digital copy, except that the News' content management system had been engineered to be difficult to log into if you weren't in the building.

As noted earlier by Anjali Mullany, who pioneered live blogging at the News and oversaw Storm Tracker, the need to establish a production process around a CMS creates a large but often hidden tax on attempts at innovation. In this particular case, the Daily News had taken a tool that could have been accessible to anyone working for the paper anywhere in the world, and added security constraints so that it instead behaved like a steam-driven printing press--workers had to be near the machine to operate it, even though the machine was a networked computer.

The defining need that drove the launch of Storm Tracker, in other words, wasn't to find new ways to inform the residents of New York City during a big storm, but simply to find a way to keep the website up when terrible engineering decisions collided with terrible weather.

This was one essential factor in the launch of Storm Tracker. There was one other. In interviews with Mullany about Storm Tracker's success, she noted that it was fortunate that Irene had hit in late August instead of early September, because in late August, most senior management were on vacation and thus could not override the decision of the News' junior but more web-savvy staff to try something new.

As noted in Section 2, institutions are designed to resist change--that is their core competence, in the language of management consultants. The risk, of course, is that too much success in that department can preserve an institution's internal logic right up to the moment it collapses. If what it takes to innovate in the manner of Storm Tracker is brain-dead technology management, the fear that your newsroom will be washed out to sea, and senior management gone fishing, then the prospects for orderly innovation among legacy organizations is grim. (As a dreadful coda, Hurricane Sandy flooded the Daily News building, and the users of the CMS suffered the same issue as during Irene. Even a year after the original crisis, no one had adapted the system to allow for a distributed workforce.)

Given this, the old news industry's collective fabulation about restoring status quo ante has itself been harmful. News organizations should obviously do what they can to improve their income, but the reliable revenue, high profits and cultural norms of the news business in the 20th century are gone, and the ecosystem that reliably produced such effects is gone as well. For individual journalists and for the institutions that serve them, cost containment, plus restructuring in the direction of more impact per hour or dollar invested, is the new norm of effective news organizations, the pattern we've taken to calling post-industrial journalism.

Post-Industrial Ecosystem

What does post-industrial journalism look like? It starts with the assumption, introduced in Section 2, that news organizations are no longer in control of the news, as it has traditionally been understood, and that the heightened degree of public agency by citizens, governments, businesses and even loosely affiliated networks is a permanent change, to which news organizations must adapt.

As one example of this change, the ejection of the Occupy Wall Street movement from New York's Zuccotti Park in November 2011 was broken not by the traditional press, but by the occupiers themselves, who sent word of the police action via SMS, Twitter and Facebook. More pictures and video of the event were generated by the participants than by the traditional media, in part because the overwhelming majority of available cameras were in the pockets of the occupiers and in part because the police closed the airspace above the park to news helicopters. Reporters on the scene hid their press badges because ordinary citizens had better access to the events in question than credentialed members of the press.

Similarly, the news organizations that ran leaked documents from WikiLeaks often described WikiLeaks as a source rather than as a publisher, on the rationale that WikiLeaks provided the material they were working from. This makes sense in a world where holders of important information can't publish it on their own and where publishers don't share source materials with one another. But there is no longer a right answer to the question, "Who is a publisher and who is a source?" WikiLeaks is a source that can publish globally; it is a publisher that collaborates on delivery of raw material with other publishers.

Coverage of events like #Occupy and Cablegate (as well as Tunisian uprisings, Syrian massacres, Indonesian tsunamis, Chinese train crashes and Chilean protests) simply cannot be described or explained using the old language of the pipeline. The best argument for thinking of news as an ecosystem is to help reexamine the roles institutions can play in that ecosystem.

Imagine dividing the new entities in the news ecosystem into three broad categories--individuals, crowds and machines (which is to say, both new sources of data and new ways of processing it). Individuals are newly powerful because each of them has access to a button that reads "Publish"; material can now appear and spread, borne on nothing but the wings of newly dense social networks. Crowds are powerful because media have become social, providing a substrate not just for individual consumption but also for group conversation. Kate Hanni was able to use newspaper comment sections to drive her "Airline Passengers Bill of Rights" because she had a better sense of those papers as watering holes than they had themselves. And machines are newly powerful because the explosion of data and analytic methods opens whole new vistas of analysis, as with lexical and social network analyses that followed the release of State Department cables.

As with the inability to make WikiLeaks stay firmly in the category of either source or publisher, there is no stable attitude that a news outlet can take toward the new agency of individuals, the spread of ridiculously easy group-forming, or the increase in the volume of raw data and the new power of analytic tools. As the Daily News' unwitting experiment with disaster blogging demonstrates, these are not resources that can be added to the old system to improve it. These are resources that change any institution that adopts them.

Now imagine dividing up the core operation of a news organization into three overlapping phases--gathering information about a story, shaping it into something ready to publish, and then publishing it. This taxonomy of a news pipeline into getting, making, and telling is of course simplistic, but it captures the basic logic of news production--take material broadly from the outside world, shape it into whatever your organization considers a story or a segment or a post, and then send the newly fashioned material back out into the world.

Armed with these two triads, we can ask, "How do individuals, crowds and machines affect the work of getting, shaping and telling?"

* As one example, the "getting" phase of the news story was the cycling blog, NY Velocity, founded in 2004 by cycling enthusiasts Andy Shen, Alex Ostroy and Dan Schmalz. Though the site existed mostly to cover bike racing in New York, the people running it grew increasingly alarmed at what they thought was a culture of willful blindness around the possibility that Lance Armstrong, seven-time winner of the Tour de France, had been doping with Erythropoietin, or EPO, a blood-boosting drug. NY Velocity interviewed Michael Ashenden, the Australian physician who had developed a test for EPO; in the interview, Dr. Ashenden went on the record as saying he believed, after testing a sample of Armstrong's blood during his 1999 Tour de France win, that Armstrong had been doping. This was original, long-form reporting, and the resulting 13,000-word interview became a public rallying point for cyclists who believed not merely that Armstrong had cheated his way to those wins, but that the professional sports journalism world was far too willing to look the other way. NY Velocity's founders were willing to pursue a lead tenaciously and publicly; not only were they completely vindicated in their suspicions, but they also demonstrated that professional journalists may simply not be covering a story well enough and that dedicated and knowledgeable insiders can sometimes fill this gap.

* To take another intersection of traditional practice and new capability, consider the way the ability to assemble groups has changed creating a story. Huffington Post's 2008 reporting project was able to cover every site of the Iowa caucuses because it could dispatch a volunteer to each site for an hour or two, something that would have been too expensive with freelancers and too travel-intensive for full-time staff. The volunteers for Off the Bus were not the people creating the eventual report on the caucuses--the project was instead a hybrid of distributed reporting and centralized writing of the story; it was, in a way, a return to the old separation of reporters in the field and rewrite men in offices close to the machine.

* Still another cross-tab of existing jobs and new resources is the way a story can be told by letting machines do some of the telling. Several projects using Ushahidi, the "crisis mapping" tool, have crossed over from "resource for recovering from a crisis" to "resource for understanding the crisis as it happens." Ushahidi has been used to create real-time maps of voter intimidation, street harassment, radiation levels and snow removal--every instance of Ushahidi for newsworthy events is an example of machines altering how data are collected, collated and displayed. Every one of the core activities of getting, making and telling is being altered by new ways of involving in dividuals, groups and machines. As noted in Section 2, the significance and pervasiveness of these alterations is likely to defeat institutions' ability to integrate change slowly. Many of the recommendations in this section are thus echoes of those from the section on institutions; when they are repeated here, it is with greater emphasis on the way that using these new resources and capabilities means adaptation to an altered ecosystem.

News as an Import-Export Business

One way to think about ecosystems is to ask what flows between its participants. As noted, flows in the 20th century were relatively linear and predictable; where there was significant complexity in flows of information, they tended to be embedded in highly specified business dealings, as with the use of syndicated or wire service copy.

The value of an Associated Press story to an individual newspaper was reflected in the interest of the locals; a subscription to the AP was justified when the value of that interest helped the paper generate more in ad revenue than the feed cost them.

This was a system where flows of business value were specified in bilateral agreements and priced in dollars--a newspaper signs an agreement with the AP in return for access to its feed. Compare that to the Huffington Post's original model: the realization that some of HuffPo's published material could excerpt existing stories, add commentary, and produce an economically viable new product.

Fair use has existed in this form for decades; what changed was the conditions of the ecosystem. HuffPo management realized that fair use, as applied on the web, meant that, in essence, everything is a wire service and that excerpting and commenting on unique content from the Washington Post or the New York Times was actually more valuable to readers than contracting with the AP or Thomson Reuters.

The Huffington Post has often been criticized for this stance, but this is shooting the messenger--what it did was to understand how existing law and new technology intersected. The AP itself is experimenting with holding back key stories from its subscribers, in a bid to acquire more direct traffic. Similarly, the AP's case against Shepard Fairey, an artist who created an iconic image of Barack Obama as a derivative work from an AP image hinged on the idea that AP had the right to photograph Obama without his permission but that Fairey couldn't use that likeness to create a related one. In the Fairey case, there was no objective reality that the case could be adjudicated on--there was simply a set of legal doctrines.

The old ethic was described by Terry Heaton in a post entitled "Why don't we trust the press?":
   Nobody ever mentions anybody else in the world of news gathering
   unless a copyright claim forces it. Before the Web, this was
   understandable, because as far as anybody knew, our reporters had
   all the angles on everything. The idea that the guy across town had
   it first was irrelevant, so why mention it? As far as our viewers
   or readers were concerned, we were the font of all knowledge.
   Besides, we had the time to gather everything we needed anyway. It
   was the world of the "finished" news product.

      But now, with news in real time, everybody can clearly see
   stories develop across all sources. We know who got it first. We
   know when something is exclusive. Our hype is just nonsense.


It has become obvious, in the new news ecosystem, that the notion of everyone producing a finished product from scratch is simply not the normal case. We are each other's externalities. This has always been the case to some degree--newspapers famously helped set the agenda for broadcast media in the 20th century--but it was often hidden in the manner Heaton describes. The explosion of sources and the collapse in cost for access has made the networked aspect of news more salient. The tech site Slashdot was clearly a source of story ideas for the New York Times' Science Times section; Boing Boing sends traffic to obscure but interesting websites, which often become fodder for stories, elsewhere, and so on.

In some ways, the ecosystemic aggregation, inspiration, excerpting and even wholesale ripping-off of journalistic content marks a return to earlier ages of newsgathering in which country newspapers often consisted of little more than week-old stories copied from metropolitan dailies. The ability to aggregate news, 18th-century style, was due in part of a lack of institutional norms (was reprinting news "illegal"? Few editors probably thought of it in those terms) and in part due to technology (few people in New York City would ever see a newspaper in rural Kentucky). The idea that news could be syndicated, for fee, is a relatively new concept in journalistic history.

The syndication model that existed under the 20th-century news production regime thus isn't coming under pressure because of bad actors, but because the basic configuration of the media landscape has changed dramatically. In the old model, reuse of material was either contractual (freelancers, wire services) or hidden. In the new model (old models, really), there are many forms of reuse; some are contractual, but most are not. The AP is a particularly visible case, but every news institution is going to have to position or reposition itself relative to new externalities in the ecosystem.

The spectrum of the exchange of value between individuals and organizations is enormous and highly gradiated--there is now an institutional imperative to get good at developing partnerships, formal and informal, that have become possible in the new ecosystem. To take one recent example, important both in itself and for what it says about the changing world, the ability to translate written and spoken material has become dramatically easier and cheaper.

Automated translation tools are far better today than they were even five years ago, as with the use of Google Translate by English speakers to read Arabic tweets; crowdsourced translation, as with dotSub or the TedTalks translators, can convert astonishing amounts of material in short periods; and the rise of institutions given to consistently bridging linguistic and cultural gaps, as with Meedan or ChinaSmack. Every institution in the world now faces two strategic choices: when, and out of what languages, do we begin translating primary course material or existing reporting to present to our audience, and, second, when and into what languages do we translate our own material to attempt to reach new audiences.

Imagining news as a linguistic import-export business, investing in importing from Arabic into English, at potentially all levels of the cost-quality trade-off, could be valuable for any U.S. newsroom that wants to cover geopolitics, while, given the demographic trends in the United States, investment in exporting from English to Spanish could add huge value in audience acquisition and retention.
Recommendation: Get Good at Working with Partners

There is a famous photo, from the 2008 Olympics, of a phalanx of
sports photographers on a platform, all angling to get what is
essentially the identical shot of Michael Phelps. The redundancy
pictured is staggering. There is something like half a million
dollars' worth of gear committed to capturing a single point of
view, and worse is the human cost of dozens of talented
photojournalists competing for minimal incremental value.

This sort of competition, where every institution has to cover the
same thing in only slightly different ways, was absurd even when
those organizations were flush. Now, with many resources gone and
more going, it is also damaging.


News institutions need to get better at partnering with individuals, organizations, even loose networks, both to expand their purview and reduce their costs. Successful examples range from the New York Times/WNYC SchoolBook partnership, designed to improve education coverage to both participants, to the aforementioned WikiLeaks and Dollars for Docs examples, to arm's length use of online data hosted by the Sunlight Foundation or Data.gov. In particular, finding ways to use and acknowledge the work of such partners without needing to fit everything into a "source or vendor" category would expand the range of possible collaborations.
Recommendation: Figure Out How to Use Work Systematized by Others

This is a subset of the previous recommendation. We are seeing a
huge increase in structured data (data that come in a highly
ordered and well-described fashion, such as a database), and a
related increase in APIs (application programming interfaces, a
systematic form of machine-to-machine conversation). Taken
together, this means a potential rise in collaboration without
cooperation, where a news outlet builds on data or interfaces made
available elsewhere, without needing to ask the institution hosting
the data for help or permission.

This is obviously valuable, as it provides low-cost, high-quality
access to previously unavailable source material. As with so many
new capabilities in the current environment, however, structured
data and API access are not new tools for doing things the old way.
These are tools whose adoption alters the organization that uses
them.

The most obvious obstacles to taking advantage of work systematized
by others are the technical skills and outlook required to use it.
This problem, fortunately, is getting somewhat better, as tools
like Many Eyes and Fusion Tables are making it easier for less
tech-savvy people to explore large data sets looking for patterns.
Even with this improvement, however, there is a need for basic
numeracy among journalists, something we've taken to calling the
"Final Cut vs. Excel" problem, where journalism schools are more
likely to teach tools related to basic video production than to
basic data exploration.

This emphasis on tools for presentation over investigation is most
acutely a problem in the nation's journalism schools, of course,
but it is widespread in the industry. (As Bethany McLean of Vanity
Fair said to us, "Anyone who's good at understanding corporate
balance sheets is likelier to work on Wall Street than cover it.")

The subtler obstacles are cultural--using work systematized by
others requires overcoming Not Invented Here syndrome and accepting
that a higher degree of integration with outside organizations will
be necessary to take advantage of new sources of data. Another
obstacle is cultural--data and APIs are often freely available, but
the hosting organizations want credit for helping to create
something of value. This imperative pushes against the
aforementioned tendency not to credit others publicly.

This logic is not just about using others' work, of course. News
organizations should do a better job of making their work
systematically available to other organizations for reuse, whether
by sharing data or by sharing tools and techniques. There will
always be a tension between competitive and cooperative logic in
the news ecosystem, but in the current environment, the cost of not
undertaking shared effort has gone up, the cost of lightweight
collaboration has gone down considerably, and the value of working
alone has fallen.

As noted in Section 2, presence of process is often a greater
obstacle to change than absence of resources. Taking advantage of
work systematized by others and figuring out ways of making your
work systematically useful to others are ways to do higher quality
work at lower cost, but doing so requires an organization to start
treating the newsroom like an import-export business, rather than
an industrial shop floor.


Self-definition as Competitive Advantage

There is no solution to the present crisis. One corollary is there is no stable state coming to the practice of news any time soon. We are not living through a transition from A to B (Walter Cronkite to Baratunde Thurston, say) but a transition from one to many, from a world where Cronkite could represent some central focus to a world with a riot of competing voices Thurston and Rachel Maddow and Juan Cole and Andy Carvin and Solana Larsen as a few members of a cast of millions.

We've seen this in microcosm--the transition from broadcast to cable networks on TV, or, as a less popular example, from terrestrial to satellite radio led to a shift from networks that catered to a broad swath of people to highly specific niches (Comedy Central, Food, and, on satellite radio, not just blues music but Delta blues or Chicago blues).
Recommendation: Always Link to Source Materials

Linking is the basic technological affordance of the web, the
feature that sets it apart from other forms of publishing, because
it says to the user: "If you want to see more on the topic being
discussed, you can find related material here." It is a way of
respecting the users' interests and ability to follow the story on
their own.

In the practice of news, the most basic form of linking is to
source materials. A discussion of a recent indictment should link
to the text of that indictment. A discussion of a scientific
article should link to that article. A piece about a funny video
should link to that video (or, better, embed it).

This is not sophisticated digital strategy--it is core
communicative ethics, yet it is disturbing that so many
journalistic outlets fail this basic test. At fault are the usual
cultural obstacles (as with Terry Heaton's observations about not
giving credit), ingrained habits (news desks used to be limited by
space or time constraints to excerpting source materials), and
commercial concern about sending readers elsewhere.

None of these obstacles, though, merits much sympathy. The habit of
not giving credit, while widely practiced, is plainly unethical.
The web no longer feels novel to the audience; it's well past time
for its core practice to be internalized by journalists. And
refusing to link for commercial reasons may make sense to the ad
sales department, but it should horrify anyone whose job involves
public service.

The public value of linking to source materials is so obvious, and
so easy, that organizations that refuse to do it are announcing
little more than contempt for the audience and for the ethical
norms of public communication.


The internet, of course, provides infinite potential variety, making the argument in favor of niche audiences (and niche loyalty) strong here as well. In addition, the old logic of geographic segmentation of local coverage allowed news outlets to buy wire service news or syndicated packages, secure in the knowledge that their audience wouldn't see the same content published or aired in a neighboring town. With the rise of search as an essential form of finding content, however, the average user now has access to thousands of sources for the story of the Somali pirates, the vast majority of which are drawn from the same wire service copy.

This creates a new imperative for news organizations, for which the strategy of "We are all things to all people in a 30-mile radius" is no longer effective. There are useful services to be rendered by hyperlocal organizations (the St. Louis Beacon, the Broward Bulldog), others by hyperglobal ones (the New York Times, the BBC), others still by highly specialized niche sites of analysis (Naked Capitalism, ScienceBlogs), and so on.

This is a breadth vs. depth trade-off. The web creates a huge increase in diversity over a world dominated by broadcast and print media. More recently, an increasing amount of news is flowing through social media sites, and especially Twitter and Facebook; the growing dominance of the social spread of news and commentary further erodes the ability for any one site to produce an omnibus news package.

There is a place for rapidly produced, short pieces of breaking news. There is a place for moderately quickly produced analysis of moderate length (the first draft of history). There is a place for careful, detailed analysis by insiders, for insiders. There is a place for impressionistic, long-form looks at the world far away from the daily confusion of breaking news. And so on. Not many organizations, however, can pursue more than a few of these modes effectively, and none that can do all of them for all subjects its audience cares about.

This is partly because institutions always face breadth vs. depth trade-offs, but the internet has made them considerably worse--masses are bigger, as with the spread of the news of Michael Jackson's death. Niches are nichier coverage of mortgage issues at Lenderama, or Latino youth issues at Borderzine. The fastest news can be faster--the White House announcement of Osama bin Laden's death was prefigured on Twitter more than once by independent sources.
Recommendation: Give Up on Trying to Keep Brand Imprimatur while
Hollowing Out Product

This is principally a negative recommendation.

Two things that have changed dramatically in the past decade are
the value of reputation (higher) and the cost of production
(lower). So many sources of news are now available that any
publication with a reputation for accuracy, probity or rigor has an
advantage over the run-of-the-mill competition. However, digital
tools have also dramatically reduced the cost of finding and
publishing information, leading to a profusion of outlets that
publish by the ton.

It is tempting for the publications with the good reputations to
combine these two changes, to find some way to extend their
reputation for high quality over new low-cost, high-volume efforts.
This was the rationale that led to the creation of the Washington
Post's blogPost aggregation and commentary feature, made famous by
the resignation of Elizabeth Flock after being reprimanded for not
having attributed some of the material she was aggregating.

It's worth quoting from the column the Post's ombudsman, Patrick B.
Pexton, wrote after she resigned:

   Flock resigned voluntarily. She said that the [two] mistakes were
   hers. She said it was only a matter of time before she made a third
   one; the pressures were just too great.

   But The Post failed her as much as she failed The Post. I spoke
   with several young bloggers at The Post this week, and some who
   have left in recent months, and they had the same critique.

   They said that they felt as if they were out there alone in digital
   land, under high pressure to get web hits, with no training, little
   guidance or mentoring and sparse editing. Guidelines for
   aggregating stories are almost nonexistent, they said.

Flock and her fellow aggregators were caught between the commodity
news logic of an aggregation site and the Post's brand, a tension
that also showed up in the New Yorker providing a platform for
Jonah Lehrer's recycled content; as Julie Bosman noted in the New
York Times, the magazine's "famed fact-checking department is
geared toward print, not the web." It also appeared in the
Journatic scandal, where fake bylines were added to stories written
by overseas freelancers.

In all of these cases, the temptation is to place a low-cost
process under a high-value brand. It's clear that rapid
commodification of ordinary news is not just inevitable but
desirable, to free up resources for more complex work elsewhere.
It's also clear that the temptation to make commodity news look
like its non-commodified counterpart is also significant, even for
institutions as august as the Post and the New Yorker.

Basic respect for the journalistic effort demands that you give
people doing commodity work clear guidelines about what is and
isn't permissible. Basic respect for your audience demands that it
be given clear guidelines about the source and process of news.
"Breaking news from around the web" can be a valuable feature and
asking people in the Philippines to write what is essentially
standard copy, given a particular set of facts, are both useful
strategies. But presenting them as no different from more
aggressively researched, composed and checked stories creates both
short- and long-term risks that are not worth the momentary
arbitrage opportunity of marrying a good brand with cheap content.


The change in the ecosystem here is that functions previously executed among competitive news organizations, and especially scoops and breaking news, are now taken over by platforms. Any given news organization may set itself up to be faster at breaking sports news than Deadspin, say, or faster at breaking tech news than Scobleizer, but no organization today can consistently beat Facebook or Twitter on speed or spread.

One final observation: A core thesis of this paper is that the country's news organizations are no longer adequate to ensuring coverage of the news on their own. This puts existing institutions in the awkward spot of needing to defend or even improve parts of the current ecosystem from which they may not profit, and which may benefit their competitors.

Were news organizations merely commercial entities, this would be impossible--Best Buy has little interest in improving the electronic ecosystem in ways that might benefit Amazon or Wal-Mart. News organizations, however, are not merely commercial entities. They are instead constituted to shield newsroom employees from most of the business questions a paper faces (however imperfect such "Chinese walls" turn out to be in practice). Indeed, if news organizations were not sources of such tremendous civic value, separate from the logic of the market, their commercial senescence would make no more difference than the closing of the local travel agent's office.

Given this, and given the need for post-industrial journalism that makes considerably better use of an hour of a journalist's time or a dollar of an institution's money, news institutions large and small, commercial and for-profit, executional and educational, should commit themselves to two changes in the current ecosystem.
Recommendation: Demand that Businesses and Governments Release
Their Data Cleanly

The most valuable dollar a news organization can make is the dollar
it doesn't have to spend, and in the 21st century, the easiest
dollar not to spend is the dollar spent gathering data. In keeping
with our recommendation that news organizations should shift some
of their priorities from covering secrets to covering mysteries,
anyone who deals with governments or businesses should demand that
publicly relevant data be released in a timely, interpretable and
accessible way.

Timely means that the data should be made available soon after
being created. It is of far less value to know what committee
recommendations were after a piece of legislation has gone up for a
vote. Interpretable data come in a structured and usable format.
Data should be made available in flexible formats, such as XML, and
not inflexible ones, like PDF. (Indeed, using a format like PDF for
publishing is often a clue that an organization has something to
hide.) Accessible means that the data are made readily available
over the public internet, instead of being kept on paper or made
available by request only. The FCC's ruling that broadcast outlets
had to publish their political ad records online, rather than
keeping them available "for inspection" at the station, was a big
improvement in accessibility.

Every news outlets should commit some resources, however small, to
taking an activist stance on this issue. Better access to better
data is one of the few things that would be an obvious improvement
for the news ecosystem, one where the principal obstacle is not
cost but inertia, and one where the news organizations' advantage
in creating improvement is not expenditure of resources but moral
suasion.

Recommendation: Recognize and Reward Collaboration

Organizations that offer grants and rewards provide a signaling
mechanism for how practitioners of journalism should regard
themselves and their peers.

These organizations should begin offering grants or create award
criteria or categories that reward collaboration, either explicit,
as in the case of SchoolBook, or implicit, as with organizations
that provide access to their data for reuse by other organizations,
as with Dollars for Docs.

Similarly, awards for successful reuse of a reporting template--for
example, other news organizations ferreting out Bell, Calif.-style
corruption--would help alter the current valorization of
handcrafted work that tends not to be repeatable, even when the
reporting uncovers a potentially widespread problem. It was a huge
loss for the nation that no organization undertook a systematic
look at other states' nursing boards after the California scandal
or a harder look for off-balance-sheet vehicles after Bethany
McLean wrote about Enron.

McLean noted, in an interview for this paper, that a key part of
her ability to study Enron was cultivating skeptics as sources--her
initial interest came after a short seller characterized Enron's
financial statements as incomprehensible. This might seem like an
obvious strategy, but few in the business press followed it, either
before the fall of Enron or, far more alarmingly, even afterward.

Organizations that shape assumed community norms among journalists
and editors should highlight efforts that build on previous work.
As with all grants and awards, these changes will reach only a few
institutions directly but will reach many indirectly, by
communicating the kinds of work that might reap either commercially
unconstrained funds, the admiration of one's peers, or both.


Conclusion

Tectonic Shifts

It was a memo from the future, an astonishing look at the dawn of the public digital landscape, from senior newspaper management. In 1992, Robert Kaiser, the managing editor of the Washington Post, attended a meeting in Japan populated by visionary technology leaders who introduced him to the future of "multimedia" and to the idea of personal computers and digital networks as alternate methods of delivery for media businesses.

Kaiser then wrote a 2,700-word memo for Post Co. CEO Don Graham and the newspaper's senior management that opened with the (inaccurate but evocative) metaphor of the frog in boiling water:
   Alan Kay, sometimes described as the intellectual forefather of the
   personal computer, offered a cautionary analogy that seemed to
   apply to us. It involves the common frog. You can put a frog in a
   pot of water and slowly raise the temperature under the pot until
   it boils, but the frog will never jump. Its nervous system cannot
   detect slight changes in temperature.

      The Post is not in a pot of water, and we're smarter than the
   average frog. But we do find ourselves swimming in an electronic
   sea where we could eventually be devoured--or ignored as an
   unnecessary anachronism. Our goal, obviously, is to avoid getting
   boiled as the electronic revolution continues.


Kaiser goes on to describe what he learned at that meeting, about a world where electronic distribution and consumption reshape the media landscape. Kaiser not only warns his fellow executives about the risks of being devoured or, worse, ignored, but he also goes on to propose that the Post immediately undertake two R&D projects: "1) Design the electronic classifieds now" and "2) Design the world's first electronic newspaper."

When the full copy of Kaiser's memo circulated among the news cognoscenti in the summer of 2012, it kicked off a flurry of public discussion about how prescient Kaiser had been, and how unfortunate it was that this incredible preview of what was coming--written before the public launch of the web--didn't get acted on.

Much of that conversation about what might have been, however, missed a second, critical aspect of the memo: Even if the Post had executed swiftly on everything Kaiser had proposed, it wouldn't have worked. Despite Kaiser's brilliance in laying out the great forces then only barely visible, his memo also contains hints of the difficulties adapting to a world where the internet was normal.

Kaiser assures his fellow executives that since people will need filters for all the new information, they will therefore need professional editors:
   Confronted by the information glut of the modern world, I suspect
   even the computer-comfortable citizens of the 21st century will
   still be eager to take advantage of reporters and editors who offer
   to sort through the glut intelligently and seek to make sense of it
   for them. Interestingly, when I asked a number of people at the
   conference what they'd like to be able to do in the electronic
   future, many spoke of finding all the extant journalism on subjects
   of interest to them. (CompuServe now offers a rather primitive
   grazing tool to permit this sort of thing.)


Kaiser looked right at the "rather primitive" capability--search--that would eventually float first Yahoo and then Google and assumed that it would remain marginal, because he assumed the business he was in--editorial judgment--couldn't be displaced. Likewise, his pair of proposed R&D efforts contained the very thinking that would sidetrack a thousand attempts at innovation; Kaiser said, of his electronic classifieds, that the Post should also
   ... reserve the right to postpone implementation until a moment
   when we're confident we'll make more money (or deter a competitor)
   by launching the electronic product.


Even someone who had seen far into the future still missed the crucial lesson, one that Alan Kay and his cohort had clearly tried to impart: No one could reserve the right to postpone implementation of the future. The big but hidden mistake was the assumption that the Post, or indeed any institution, could opt out of the coming changes. This mistake was made more punishing because Kaiser's assumptions didn't allow for the possibility that new distribution channels for news and advertising might generate less money per user, rather than more.

This was the real issue, impossible to recognize then, but obvious in hindsight: The problem legacy news organizations faced over the two decades since Kaiser's trip wasn't competition but revolution. They assumed that new technology would raise rather than lower ad revenue, or that it would deliver more control to the publisher rather than to the reader. This was consonant with everything that had happened up to 1992, but it wasn't what was about to happen as the internet started giving everyone a lot more freedom.

Tectonic Shifts

In the 1990s, those of us thinking about the relationship between the internet and news organizations wrongly assumed that the core problem those organizations faced was understanding the future. This turned out to be a merely ancillary problem. The core problem was adapting to that future.

The story of journalism in 2012 is still often told as the story of the breakdown of the old world, the end of the period when "the news" was whatever an enumerable collection of institutionally stable actors chose to publish. This assumption ran so deep that even someone who had seen decades into the future could still believe that the digital turn in the newspaper business would favor traditional virtues of editorial choice over the new ones of user empowerment and that the business case for electronic media was around revenue generation rather than cost reduction.

That "End of an Era" story, though, is itself ending. We are living in the least diverse, least inclusive media environment we will inhabit for the foreseeable future, which is to say that the ecosystem forming around us will include more actors and actions than even today's environment does.

It's easy to equate this increase in public speech with an increase in chaos, but chaos is a wasting asset--what seems hopelessly confusing today will be tomorrow's new normal. The old order won't be restored, but people will get used to the new one that's emerging.

Though we have generally concentrated on the question "What does the production of news like today?" in this section we will ask a related question: Given the forces already at work, what will the production of news look like in 2020, seven years from now? This is as far from today as today is from 2006, when You-Tube, Twitter and Facebook were all in their infancy.

As with any exercise in prediction, we will get this at least partly wrong, overestimating some changes, underestimating others, and, most significantly, failing to predict new forces that will appear in the next seven years. Our goal here is for accuracy in direction, not endpoint; we believe that many of the forces that will shape the news environment in 2020 are already visible today, in the same way that social networking and user-distributed video were visible seven years ago.

In 2020, there will be considerable surface continuity with the news environment of the 20th century. There will still be a Los Angeles Times and a CNN. Yet this continuity of institutions will be accompanied by a reconfiguration of almost every bit of the media world in which they operate. As George W.S. Trow put it in "Within the Context of No Context," his wonderful and strange musing on the changed social landscape in the United States:
   Everyone knows, or ought to know, that there has happened under us
   a Tectonic Plate Shift [...] the political parties still have the
   same names; we still have a CBS, an NBC, and a New York Times; but
   we are not the same nation that had those things before.


Trow was discussing the loss of any obvious core of civic culture in the aftermath of the late 1960s, but the figure of a tectonic shift can also serve as a metaphor for the media environment today. The label "CBS News" still describes the journalistic arm of a U.S. broadcaster, but it no longer stands for "Tiffany" standards in news, and it no longer occupies a position of unquestioned centrality in the news environment. This is partly because CBS itself approaches news differently, but mostly because the competitive and consumptive landscape of news has shifted so dramatically that even if CBS News' sole goal in the last two decades had been to preserve its former position, it would have failed.

The news ecosystem of 2020 will be a study in expansion, with heightened contrasts between extremes. More people will consume more news from more sources. More of these sources will have a clear sense of their audience, their particular beats or their core capabilities. Fewer of these sources will be "general interest;" even when an organization aims to produce an omnibus collection of news of the day, the readers, viewers and listeners will disassemble it and distribute the parts that interest them to their various networks. An increasing amount of news will arrive via these ad hoc networks than via audiences loyal to any particular publication.

Almost every aspect of the news environment will be more variable than it is today. We're not shifting from big news organizations to small ones, or from slow reporting to fast. The dynamic range of journalism is increasing along several axes at once. The internet has unleashed demand for more narrative and more data-driven news, for a wider range of real-time sources and wider distribution of long-form pieces.

A few organizations will have larger newsrooms than today, mostly subsidized by media services sold to professionals (as with Thomson supporting Reuters, or Bloomberg's purchase of Business Week.). Most news outlets, though, will have smaller newsrooms, measured by full-time headcount. At the same time, there will be many more niche players than today, with smaller and more narrowly tailored operations (the Outer Banks Voice; Hechinger Report).

There will be more nonprofit news organizations, driven by several kinds of donation--direct cash subsidy by philanthropies and other donor organizations (Ford Foundation funding Los Angeles Times reporters; William Penn Foundation funding PennPraxis), user donations of cash (NPR; TPM), and in-kind donations of the time and talents of a particular community (as with the creation of Wikipedia disaster articles, or Twitter hashtag streams).

The obvious benefit of increased subsidy for news is its increased availability. The equally obvious downside is that it risks further blurring the boundary between public relations and journalism. The growing number of news outlets, and their varying motivations and funding sources, increases the need for self-policing, as independent news outlets learn to better identify, label and publicly rebuke "churnalism." (As David Weinberger has noted, transparency is the new objectivity.)

The decay of the traditional agenda-setting function of the press will continue, and with it the idea of "the public" as a large, interconnected mass of news-consuming citizens. Choice in available media outlets will continue to expand, leading not so much to echo chambers as to a world of many overlapping publics of varying sizes. Seen in this light, the long-term collapse of trust in the press is less a function of changing attitudes toward mainstream media outlets than a side effect of the continuing fragmentation of the American media landscape. (It is probably time to retire the idea that there is something called "the press" that enjoys a reputation among some group called "the public.")

The shift in control of distribution will also continue. The old model, where most users visited a home page or used a mobile application tied to a single organization, will continue to lose ground to superdistribution, users forwarding relevant materials to one another. We already live in a world where the most widely circulated stories acquire audiences that dwarf the median headcount. Adapting to this increasingly unequal distribution will require that most organizations get better at working with their users to filter and pass on relevant material.

This superdistribution won't just be about the spread of new material; one of the great surprises of Twitter, a medium built around "short" and "now," is how much demand it has exposed for long-form writing and video. News.me, a recent startup, filters through people's Twitter feeds and recommends the most widely viewed links from previous 24 hours; a remarkable amount of what gets surfaced is not singing cats but long, careful pieces of reporting or opinion.

Though the "hamster wheel" (chasing transient viewers with rapid publication of sensational stories) is an obvious effect of the internet's colonization of the news landscape, the increase in the dynamic range of the news environment is taking place at both ends of the distribution; the hamster wheel has been accompanied by an increase in large-scale reporting and analysis.

More techniques will be deployed in the production of news--algorithmic data analysis, information visualization, solicitation of amateur input, feedback loops with crowd reaction, automated production of data-driven stories. More generalists will be working in niche subjects; interviewers on particular topics who create, edit and distribute photos, audio or video as a newsroom of one. Narrower and deeper specialization will occur among the newsrooms that have staffs large enough to allow collaborative units to work together: By 2020, the most expert data miner, information visualizer or interactive experience designer will have a far more refined set of tools and experience than any of those people do today.

Each newsroom will become more specialized, with less simple replaceability of employees and functions from one newsroom to the next. Each newsroom will have a better sense of who its partners are, among institutions and the general public, and will have customized its sense of how best to work with them. Many producers of the kind of material we used to regard as news won't be news organizations, in any familiar sense of the word. The police blotter will come from the police. Environmental data will be presented via interactive tools hosted by the Sierra Club. Wikipedia and Twitter will strengthen their roles as key sources of information for breaking news.

As Kaiser and the Washington Post eventually found out, there's no reserving the right to postpone implementation for the kinds of changes we are witnessing. There is only the struggle to adapt and to secure a niche in the ecosystem that allows for stable creation of long-term value.

What Should Journalists Do?

Like a Necker cube, it's possible to look at the journalism landscape and see one of two sets of relationships--the work of individual journalists supporting institutions, or the work of institutions supporting individual journalists. There is some truth to both views, of course, but we have concentrated on the latter, for several reasons.

First, the work of journalists takes both logical and temporal precedence over the work of institutions. Second, the act of witnessing, discovering or understanding what is important, and then conveying that in a way that various publics can understand, is the sacred task; concern for journalistic institutions takes on public urgency only to the degree that they help the people engaged in those tasks. And third, far too much of the conversation of the past decade has assumed that the survival of existing institutions is more important than the ability of anyone to take on that sacred task, however it's done.

Though the concept has been somewhat tainted by the cheesiness of "Brand You!" boosterism, we live in an age where the experiments of individual journalists and small groups are ideal for identifying possible new sources of value--process is a response to group dynamics, so the smaller the group, the easier it is to balance process and innovation (though later, of course, those innovations will have to be rendered boringly repeatable).

If you were looking for an ideal mantra for a journalist, writer, analyst, media artist, data miner or any of the other roles and tasks that matter today, "Proceed until apprehended" is a good one. As an NPR executive said to Andy Carvin during his invention of the curated Twitter news feed, "I don't understand what you're doing, but please keep doing it."

In this paper, we've offered a description--several, actually--of the skills and values an individual journalist can bring to bear. The range of these descriptions exists because journalism is not moving from A to B, from one stable state in postwar America to some new, alternate state today. Journalism is instead moving from one to many, from a set of roles whose description and daily patterns were coherent enough to merit one label to one where the gap between what makes Nate Silver a journalist and what makes Kevin Sites a journalist continues to widen.

With the coming increase in possible modes and tempos of journalism, our overall recommendations for journalists are these:
   Know yourself. Know what you are good at and what you are not good
   at, and know how to explain those things to others. Know your areas
   of expertise, both for content (North African politics; civil
   engineering; historical weather patterns) and skills (are you an
   interviewing journalist? A researching journalist? A Final Cut
   journalist? An Excel journalist? A Hadoop journalist?).

      Know when the tools you need are algorithms or crowds. Know when
   a person you need to talk to is more likely to be found via Twitter
   than directory assistance. Know when your network can help; know
   when someone in your network's network can help, and get good at
   asking for that help (and also get good at rewarding people who
   help).

      Know when process is aiding your work and when it's not, and, to
   the degree you can, know when to break the glass in the latter
   case. Know when to work alone, when to call for help, when to
   partner outside your usual sphere.


Much of this is about specialization of one sort or another. It's possible to specialize in content: in the kind of material you cover, or the kind of background you master, or the kind of people you interview. It's also possible to specialize in technique: you can get good at mining databases, reading investment documents, traveling in distressed zones, or engaging users, and each of those skills will be transferable to many areas of inquiry. You can specialize in content and be a generalist about technique, you can specialize in technique but be a generalist in content, or you can specialize in both. (Specializing in neither used to be a fine answer; less so today.)

Journalism schools will have to adapt to these changing models as well. Already, journalism schools are more like film schools than law schools, which is to say that the relative success or failure of a J-School grad is going to be far more variable than it used to be. There are fewer entry-level jobs the jobs that used to serve as unofficial proving grounds and apprenticeships --in metropolitan dailies and local TV than there used to be. Furthermore, the careers students head into will be more variable, and more dependent on their ability to create their own structure, as opposed to simply fitting into a position in a well-known collection of rich and stable institutions.

Schools should respond by helping students both understand what sorts of specializations they'd like to undertake, and how to go about them, a task that has much less to do with fitting them to particular institutions, as with the old (and now harmful) broadcast vs. print split, and much more to do with fitting them to particular forms of inquiry, wherever and however they practice it.

The fate of journalism in the United States is now far more squarely in the hands of individual journalists than it is of the institutions that support them. To get the kind of journalism that a complex, technocratic democracy requires, we need the individual practitioners to take on the hardest part of the task of working out what constitutes good journalism in a world with no publishing scarcity.

What Should Legacy News Organizations Do in This Environment?

Though many existing institutions still regard the principal effect of the current changes as continued loss of revenue, the restructuring of American journalism is far more influenced today by organizational models than by income (or lack thereof). With a handful of exceptions, for-profit news organizations will have to continue to reduce expenses to below still-falling revenue, but simply cutting will leave us with legacy institutions that do less with less.

Existing institutions must adapt their journalistic operations, not just their bottom line, to the internet. Doing more with less is always easier said than done, but as Homicide Watch or Narrative Science demonstrate, it is not impossible.

Though we put forward several recommendations in the body of this paper, our overall recommendations for legacy institutions are essentially these:
   Decide what part of the news you want to report and how. Get out of
   any activity that doesn't support those goals. Look for
   partnerships or collaborations with other organizations that
   advance those goals at lower cost than you could manage in-house.
   Work to make the remaining activities either excellent or cheap
   (or, ideally, both).


Some legacy news organizations will simply shrink the cost of filling the news hole with no other reorganization, a move that will amount to getting out of the hard news business by degrees. Some of these organizations may be able to survive with their newly lowered expenses, but the reason to care about the continued health of legacy news organizations has always been about the public service they provide; organizations that shrink without trying to take on new, cheaper capabilities are abandoning at least part of that public service mission. They will also attract fewer competent journalists.

Keeping expenses below revenue remains a problem, of course. Advertising declines--six years and counting--have left the nations' newsrooms, subsidized by that money, in a parlous state. Given advertisers' continued decampment to alternate platforms and the dreadful logic of a declining print audience--income compresses faster than costs for running the presses many legacy organizations will have to operate with an expanded sense of where revenue can come from: running events, applying for grants for specific beats, digital membership dollars from the most committed 5 percent of readers. Continued reduction of costs, however, remains the most obvious strategy.

There is no way to support the old "one-stop shop" model for supplying all (or even most) of a user's news and information, because, without geographic barriers to entry, there is very little defensible advantage in running commodity news that's the same as in the next town or state over. Like the principle of subsidiarity for the U.S. government (that the federal government should ideally run only those services not better run by the states, states than cities, and so on), news should be produced and distributed by the people best able to cover it. This suggests a shift to dramatically increased specialization and partnership.

In practice, many legacy newspapers have followed this advice by loading their front pages with reams of AP content and the occasional "big blowout" news story, a prime example of adapting to loss of income rather than adapting to the internet. A digitally minded news organization would dispense with running commodity content online entirely, perhaps linking to important news, or even excerpting the content of smart bloggers or other aggregators. No matter what specific decisions get made in this regard, however, news institutions that see the "front page" as their primary organizational concern will miss many opportunities for reinvention.

The wastefulness of pack journalism and the empty calories of unimproved wire service news are both bad fits for most institutions in the current environment. The organizations that set out to provide a public with a large part of the news will more often be aggregators, in the manner of Huffington Post or BuzzFeed, than traditional news organizations, if only because the cost and quality curve favors that form of aggregation over expensive improvements of syndicated content or, further up that curve, creation of custom material that lacks either a passionate audience or a long shelf life.

Similarly, newsrooms will have to decide what parts of their operations to commodify. Much checklist reporting (e.g., brief pieces on last night's game or this quarter's sales figures that have to be present, but don't have to be long or excellent) can be replaced by aggregation, or by machine production. For most organizations, anything that is high touch but low value (and by high touch we mean anything that involves more than 10 minutes of paid human attention) should be automated, outsourced to partners or users, or removed entirely.

Newsrooms that have mixed reporting goals--breaking news and long-form analysis--will have to get better at understanding the sharpening tradeoffs between speed and depth. There is no right answer here, or even right mix: Coverage of slow-moving beats with a relative handful of relevant participants--the mining industry, or automotive design--will have a different mix than fast-moving, surprise-driven ones--electoral politics, civil wars.

Similarly, newsrooms will have to understand the sharpening trade-offs between aggregation and original reporting, and optimize for each activity differently, or the trade-offs between translating first-person accounts vs. putting journalists in between those sources and the audience, to contextualize and interpret.

Existing organizations will also have to get better at managing both relationships and data as new resources. The ability of an institution to ask its own users to participate in the creation, vetting and distribution of news, or to find firsthand witnesses or knowledgeable insiders for a particular story, will become key sources of differentiation. Similarly, the ability to master certain sorts of data and to be able to reliably create value from it over time are increasingly essential skills. (The irony of U.S. News and World Report's long competition with Newsweek and Time is that the U.S. News college listings database may soon become more valuable than those other two publications combined.)

On the procedural front, organizations will have to get better at understanding when process is a help and when it is a hindrance, and they will have to get better at making their process "hackable." They will likewise have to decide which employees or volunteers have the ability to violate or alter the standard institutional processes, in order to pursue unforeseeable but valuable opportunities. Of all our recommendations, this one may be the most difficult for legacy institutions to follow. But in other ways, the success or failure of many of these companies will be determined by their ability to embrace flexibility.

What Should New News Organizations Do?

The range of new models and ideas being tried by journalism startups is wide, but most of the groups working on these ideas are not yet either robust or stable. This is partly because, as in any revolution, the old things break long before the new things are put in place, but it is also because the business model in the last several decades has created a news monoculture, where advertising subsidy has been the default revenue source even for those organizations that also charged fees directly to their users.

New news organizations will have to do everything legacy organizations do in terms of mastering the trade-offs between speed and depth, aggregation vs. origination, or solo creation vs. partnership. In general, however, understanding and managing these trade-offs is easier for new organizations, simply because individual employees don't have to unlearn previous assumptions in order to adapt to present realities. As always, the advantage young people and organizations have over older ones isn't that they know more things. They don't. The advantage is that they know fewer things that aren't true any longer. Without carrying the weight of accumulated but maladaptive assumptions, they have to spend less time and energy unlearning things before they can see and respond to the present world.

Our overall recommendation for new news organizations is even simpler than for journalists or for legacy organizations:

Survive.

The visible crisis of news institutions is the shrinking of their traditional functions, but the second, less discussed, crisis is the need for institutional stability, predictability and slack among the nation's news startups.

Much of the question of institutionalization of startups concerns the way these organizations manage income and expenses, a conversation outside the scope of what 21st-century journalism looks like. (To reiterate our position: Most of the for-profit vs. nonprofit discussion is useless. Any way of keeping expenses below revenue is a good way.) Some of it, however, has to do with organizational assumptions and capabilities built into new organizations from the start.

New organizations should assume that cost control is the central discipline and that many sources of subsidy available to startups have a limited lifespan. They should master working with amateurs, crowds, machines or other partners to keep cost low and leverage high. To survive, new news startups will need to take on some of the routinization of work and stabilization of process of the older institutions they are challenging. They should not fear becoming a little boring.

There is a certain blitheness to the conversations around the current disruption, a kind of "great cycle of life" belief that the old institutions will be weakened and that the new institutions will then automatically take their place.

That is one possible scenario, of course. Another is that the old institutions become weakened but that the new ones don't take their place, because they lack the institutional stability to act as a counterweight to large, bureaucratic organizations. Of all the terrible scenarios it is possible to imagine, this is the worst one--the legacy organizations continue to diminish in force and function, but the new entities arising simply can't be as effective a check on bureaucratic power.

The End of Solidarity

Perhaps the most salient change in the next seven years will be the continued weakening of the very idea of what constitutes news, and thus what constitutes a news organization. This change, long since begun by Jon Stewart and MTV election coverage, is still at work today; to the question "Is Facebook a news organization?" neither "yes" nor "no" is a satisfactory answer. (A better reply is "Mu," programmer-speak for "The question as asked has no sensible answer.") Facebook is critical to the news ecosystem, yet it is organized along lines that are out of synch with anything we would recognize as a journalistic organization; its presence alters the context of that question.

There will also be fading clarity as to what constitutes "the news," full stop. Institutions persistently mistake shallow continuity for deep structure; news isn't a coherent or ontologically robust category; it is a constantly negotiated set of public utterances by a shifting set of actors, one that happened to go through a period of relative stability in 20th-century America. We are seeing the end of that stability, the end of the curious bookkeeping that says that the St. Louis Post-Dispatch is a news organization, even though it runs Annie's Mailbox and the funny pages, while Little Green Footballs is not, even though Charles Johnson did a better job than CBS in vetting the phony National Guard memos involving George W. Bush's service.

The production of news has moved from being a set of jobs to a set of activities; there will always be a core of full-time practitioners, but there will be an increasing amount of participation by people working part time, often as volunteers, and distributed by people who will concentrate less on questions of what is news and what isn't than on questions like "Will my friends or followers like this?" Increasing overlap and collaboration between the full-and part-time practitioners, and between the employees and the volunteers, will be a core challenge over the rest of the decade.

This will be a world where the biggest changes have come in the roles of not full-time journalists, but of the public, where atomized consumption and private discussion in small groups have given way to a riot of alternate ways of sharing, commenting on and even helping shape or produce the news.

All of us are adapting to this changed environment; the existing institutions and the new ones, the full-time shapers of the news and the part-time ones, the generalists and the specialists. And perhaps the single most important adaptive trait is to recognize that that we are in a revolution, in its sense of a change so large that the existing structure of society can't contain it without being altered by it.

In a revolution, strategies that worked for decades may simply stop working (as many already have). Strategies that seemed impossible or insane a few years ago may now be perfectly suited to the current environment. This period is not over, and the end is not even in sight; the near future will hold more such reversals, so that even up-to-the-minute strategies of a few years ago (RSS feeds and staff blogs) may fade into prosaic capabilities, while new ones (the ability to hunt for mysteries instead of secrets, the ability to bring surprising new voices to public attention) may become newly important.

More than any one strategy or capability, the core virtue in this environment is a commitment to adapting as the old certainties break and adopting the new capabilities we can still only partially understand, and to remember that the only reason any of this matters to more than the current employees of what we used to call the news industry is that journalism--real reporting, about whatever someone somewhere doesn't want published--is an essential public good.

Methods Used in This Report

We drew on a variety of methods while formulating our analysis, recommendations and conclusions. Primarily, the research was based in qualitative interviews, conducted both one-on-one, on location, over email or telephone, and at the offices of the Columbia University Graduate School of Journalism. A significant amount of data was gathered at a closed-door conference at the journalism school on April 17-18, 2012, that involved 21 people.

For the most part, however, this paper draws on the industry experience and previous scholarship of its authors. It attempts to combine more traditional academic theory with current developments in the worlds of journalism and digital media--always a fraught task. To the degree we have succeeded, we hope that the paper is neither superficial to those coming to it as scholars, nor overly dense to working journalists who may work their way through its pages.

Ultimately, we believe that this paper should also serve as a call to further, more traditional academic research. Many of its conclusions can be tested through a variety of methods and with a variety of goals in mind. Insofar as the authors each work at different schools of journalism in New York City, and insofar as each is engaged in a different aspect of scholarly production for their respective home institutions, the future for "useful journalism research" would appear bright. Ultimately, the conclusions and provocations of this paper will rise or fall based on changes within journalism itself.

Acknowledgments

In keeping with the spirit and theme of this paper, this has been a collaborative effort, inadequately accounted for by the authors' names on the cover. All of us have benefited from observations, conversations and counsel from our colleagues who have taken time to support this effort in one way or another.

Our first thanks goes to Charles Berret, a Ph.D. candidate at Columbia Journalism School, who has been with us all along and helped both coordinate and think through the various aspects of this work. We couldn't have done this without Charles.

We thank Dean Nicholas Lemann of Columbia Journalism School, whose vision for this examination of the journalism landscape gave the project its start. None of this would have or could have happened without him. Others in the Columbia administration who aided us greatly in the process were Sue Radmer, Stephen Barbour and Anna Codrea-Rado. We thank Marcia Kramer for her patient copy editing and suggestions.

We are indebted to the support of the Carnegie Corporation, the funder of this work. We would also like to thank the Tow Foundation for its ongoing support in all our work at Columbia through the Tow Center for Digital Journalism.

The largest collection of voices represented here, and the people who gave the most time, participated in a two-day meeting and workshop, held in New York City on April 17-18, 2012. Attendees included Chris Amico, Laura Amico, Josh Benton, Will Bunch, Julian Burgess, John Keefe, Jessica Lee, Anjali Mullany, Shazna Nessa, Jim O'Shea, Maria Popova, Nadja Popovich, Anton Root, Callie Schweitzer, Zach Seward, Daniel Victor and Christopher Wink. It is no exaggeration to say that we began that meeting with a disparate set of observations and ended it with the outline of this work.

Throughout the process, we have relied on the observations of our peers, either as interviewees on the present and future state of journalism, or as respondents to early drafts of the work. For that, we thank Erica Anderson, John Borthwick, Steve Buttry, David Carr, Andy Carvin, Susan Chira, Reg Chua, Jonathan Cooper, Janine Gibson, Kristian Hammond, Mark Hansen, Andrew Heyward, Alex Howard, Vadim Lavrusik, Hilary Mason, Bethany McLean, Javaun Moradi, Dick Tofel, Matt Waite and Claire Wardle. University faculty, working inside and outside traditional journalism schools, proved to be vital sources of intellectual provocation and nourishment; we thank, in particular, Rasmus Kleis Nielsen of the Reuters Institute for the Study of Journalism at the University of Oxford and Michael Schudson and Robert Shapiro of Columbia University.

Thank you, lastly but not least, to our families for the tolerance, support and advice throughout.

C.W. ANDERSON

christopher.anderson@csi.cuny.edu

College of Staten Island,

The City University of New York

EMILY BELL

ebell@columbia.edu

Columbia University

CLAY SHIRKY

clay.shirky@nyu.edu

New York University

Received 12 August 2014 * Received in revised form 13 February 2015 Accepted 18 February 2015 * Available online 30 July 2015
COPYRIGHT 2015 Addleton Academic Publishers
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2015 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Title Annotation:p. 77-123
Author:Anderson, C.W.; Bell, Emily; Shirky, Clay
Publication:Geopolitics, History, and International Relations
Article Type:Report
Date:Jul 1, 2015
Words:22033
Previous Article:Post-industrial journalism: adapting to the present.
Next Article:Parsing TTIP's geopolitical implications.
Topics:

Terms of use | Privacy policy | Copyright © 2018 Farlex, Inc. | Feedback | For webmasters