Category Archives: press criticism

what sustains free speech?

My remarks last week at a small conference on “Tolerance, Citizenship, and the Open Society” at the Tisch College of Civic Life …

We human beings did not evolve to take a broad view of justice, to collect information from diverse sources, to reason impartially, and to be responsive to other people who differ from us. These acts do not come naturally to us.

But we are capable of building prosthetics. For instance, we did not evolve with the skill to tell time precisely, which is now useful for coordinating behavior in mass societies. So we wear wristwatches, hang clocks on our walls, and display the current time on most of our electronic devices. A clock or a watch is a prosthetic device that extends our natural capacities.

An invention (in this case, a clock) will not suffice on its own. Many people must use it. That requires some kind of system that creates incentives or requirements for producing the devices and using them widely. A market with supply and demand can work; so can a state mandate. Either one is an institution.

We have created institutions that extend our ability to deliberate about justice. An example was the metropolitan daily newspaper from ca. 1910 to ca. 1990. Always very far from perfect, it nevertheless delivered important, mind-broadening information to about 80% of Americans every day in the year 1970. They (and advertisers) paid for the local press because it also provided sports, classified ads, comics, and whole package of goods–but with the most important news on the front cover, where it could not be missed.

A university is another institution that supports inquiry and discussion about important matters. It is more complex than a newspaper. Its revenues may include tuition, government aid, grants, gifts, intellectual property transfers, and clinical fees, among other sources. The goods it produces include skills and knowledge of value to each learner; virtues and skills that have public value; the pure public goods of basic knowledge and culture; monetizable forms of knowledge, such as patents; services, such as meals, art exhibitions, and clinical care; and credentials and entry to the middle class.

The skeptical view of such institutions is that their underlying economic motivations determine the ideas and discussions that they support. For example, newspapers are owned by tycoons or faceless corporations that just want to maximize profits. Universities sell social stratification and individual advancement. This analysis always merits attention and explains some of the phenomena. But it is one-sided, because these institutions are also the result of human artisanship–of people creating the means to sustain better thinking at a large scale.

For instance, the metropolitan daily newspaper can be interpreted as the product of the media industry, but it should also be seen as the product of the press. Traditional newspapers tried to distinguish the two by separating the newsroom from the publisher’s suite, but those subsystems were connected. For instance, plenty of publishers were former shoe-leather reporters. Their motives were mixed. That is good because mixed motives produce scalable public goods.

Too simple a theory would yield two predictions about newspapers that both proved incorrect. In 1900, you might predict that millions of people would never spend their own money voluntarily to purchase relatively impartial and challenging daily news. But they did–in part because they were also buying comics and box scores. In 1970, you might predict that we would always have a press, because it meets a social need. But the press has collapsed (half as many people work as reporters today compared to ten years ago) because the Internet has killed its business model.

As with other forms of artisanship, nothing is for certain. Ingenuity, commitment, and perseverance are required. The institutional structures that support broadened understanding depend on intentional work.

The results are always flawed. The recent scandals with college admissions just bring home the flaws of universities, for instance. We should have a free, open, informed, and consequential discussion about how to improve them. But no discussion can occur outside of a viable forum that depends on an institution. We don’t spontaneously gather to discuss; the discussion always happens in a university or a school, an op-ed page of a privately-owned newspaper, Facebook, a union hall, a church basement, a party convention, the state legislature–somewhere that draws resources and assembles users.

These institutions then structure and limit the discussion. There is no view from nowhere, only a permanent struggle to discuss as wisely as we can in various forums. We don’t create these forums deliberatively; most of them arise as the result of accident, power, or leadership. Because they are all flawed and limited, it is essential to have many of them, with diverse forms, competing and checking one another.

This is a “civic” perspective because it emphasizes our ability to shape the world of discourse through artisanship. And it broadens our attention so that we consider not only the rules for speech within an institution (e.g., campus speech codes) but also–and usually more importantly–the underpinnings of the institution itself.

See also a civic approach to free speech; Sinclair and Bezos: media ownership and media bias; don’t let the behavioral revolution make you fatalistic; prospects for civic media after 2016; China teaches the value of political pluralism; polycentricity: the case for a (very) mixed economy.

the Historovox

Corey Robin’s essay, “Why Has It Taken Us So Long to See Trump’s Weakness?,” is mainly interesting as an argument about trends in reporting. Robin criticizes

a new genre of journalism that forgoes the pedestrian task of reporting the news in favor of explaining it through the lens of academic research. Ensconced at Vox, FiveThirtyEight, dedicated pages of the Washington Post and the New York Times, and across Twitter, the explainers place great stock in the authority of scholarship — and in journalists who know how to wield the authority of scholars

He argues, “There’s a bad synergy at work in the Historovox — as I call this complex of scholars and journalists — between the short-termism of the news cycle and the longue durée-ism of the academy. … When academic knowledge is on tap for the media, the result is not a fusion of the best of academia and the best of journalism but the worst of both worlds.”

An obvious objection is that there are more than just two genres of writing about politics (academic analysis and pedestrian journalism). I’d place Robin’s “Historovox” on a longer list.

  1. Old-school deadline-driven political reporting: The writer tells you what happened yesterday. The lede is an event: a speech, an endorsement, a vote, an indictment. Subsequent paragraphs tell (or remind) you what happened earlier, leading up to this new event. To the extent that the news is explained, the available explanations include: what the actors and their spokespeople say happened, how their critics reply, and the tactical advantages that will result for each. An imaginary example: “The Senator traveled to Wisconsin today to talk about jobs. This follows on the heels of her speech about the environment in Los Angeles last week. People involved in her campaign said that she is engaging two important constituencies. Her opponent charged that she wants ‘to rake in the dollars from spoiled Hollywood liberals.’ Of course, prospective presidential candidates always test their support in key states.”
  2. Positivist, mostly quantitative academic scholarship: The writer looks for statistically significant patterns in representative samples of data (rather than “anecdotes”). She poses and tests explicit hypotheses. She situates her original results in the context of peer-reviewed literature. For instance, “Some previous studies suggest that candidates mainly appeal to donors. Other studies suggest that they focus on ordinary voters. Our analysis of 256 campaign events finds that donor-appeal explains 11% more of the variance in decisions about candidate travel.”
  3. Ideological advocacy: The writer hopes to advance conservatism, or socialism, or environmentalism, or whatever, and uses recent political events as evidence and as a “hook” to persuade the unconvinced and mobilize the base. “The Senator made a great speech about jobs in Wisconsin but needs to remember why unions have declined. It’s no accident that wages have fallen as union membership has fallen: these are the results of neoliberal policy choices.” This style extends from opinion magazines and op-ed pieces deeply into academic journals.
  4. Theory-building: The writer is primarily interested in developing and defending general social theories, which may have both normative and explanatory components. She is trying to develop, for example, a new version of civic republicanism or intersectional feminism or social capital theory. As in #3, recent examples serve as illustrations and “hooks,” but the argument is less predictable, less topical, and may be considerably more complex.
  5. The “Historovox” is a fusion of #1 with #2 and/or #4. Its typical style is to “explain” a concrete recent event by summarizing some relevant positivist social science (#2) and adding an interesting social theory (#4). The very bright, broadly-educated reporter works by searching the scholarly literature and interviewing academics. This style claims to avoid #3, which is seen as politically biased, in favor of “research.”

Robin offers a subtle defense of #1–traditional deadline journalism–by way of quotations from political theorists who might be seen as “particularists”: highly skeptical of generalization and concerned with attending to details:

Everyone knows and cites Orwell’s famous adage: “To see what is front of one’s nose needs a constant struggle.” Less cited is what follows: “One thing that helps toward it is to keep a diary, or, at any rate, to keep some kind of record of one’s opinions about important events. Otherwise, when some particularly absurd belief is exploded by events, one may simply forget that one ever held it.”


the job of the scholar is to recall and retrieve what the Marxist critic Walter Benjamin described as “every image of the past that is not recognized by the present.” The task is not to provide useful knowledge to the present; it is to insist on, to keep a record of, the most seemingly useless counter-knowledge from the past — for the sake of an as-yet-to-be imagined future.

If I read him right, Robin is not saying that we need writers to express opinions about each event and record what they have opined. Rather, we–readers, citizens–should do that. Journalism gives us the raw material for our daily opinion-formation, and we should hold ourselves accountable by checking our views as new data arrive.

As a bit of a particularist myself, I find these quotes resonant, and I start with the premise that we badly need paid professional reporters to cover events. But the objection to #1 is that it was never theory-free, never just a record of what happened yesterday. Instead, it always embodied a problematic general theory, according to which history results from explicit decisions by self-interested professional politicians who compete with each other. Absent are deeper causes, issues ignored by the major parties, areas of agreement, and the work of citizens. Thus deadline journalism never served citizen-readers as well as it should have. It served up the wrong mix of “news events” for us to form opinions about.

#2 is valuable but has its limitations. As I argued right after the 2016 election, positivist social scientists mostly failed to predict Trump because their job is to detect trends in data collected already (i.e., the past). They can’t see that something is about to shift fundamentally, and when that happens, they retain a bias in favor of treating the new event as one outlying datapoint that doesn’t threaten the theory. A classic version of that critique is Robert C. Lieberman, “Ideas, Institutions, and Political Order: Explaining Political Change.” American Political Science Review 96.4 (2002): 697-712.

Another problem with #2 is tempo. The process of collecting representative data, analyzing it, and publishing it in peer-reviewed form takes many months or years, by which time events have moved on. Citizens cannot benefit from analysis unless they can use it in time.

Everyone criticizes #3–editorializing in support of an ideology–yet ideologies are indispensable heuristics, and each case of advocacy can contribute to a rich public sphere as long as you read it critically along with other views.

The advantage of #5 is translation. It connects social theory and empirical data to the news, allowing readers/citizens to learn from scholarly expertise. The big disadvantage is that there are theories for every fact. As Robin observes, when Trump looks strong, it’s time to cite the literature on authoritarianism. When he’s weak, we dust off the literature on the weak presidency. Historovox writers have a Malcolm-Gladwellish tendency to discover a new idea and find evidence of it everywhere for a while. Then events change, interests wander, and they find a new idea. As he argues, this is no way to learn.

But I think several commentators on Crooked Timber are right that explanatory journalism strives to address a real need. If we only had the first four categories listed above as separate streams, we’d be crying out for linkages. Sites like Vox and FiveThirtyEight (and The Conversation) don’t do this perfectly, but they seem fairly self-reflective and dedicated to self-improvement, and nobody could pull it off perfectly at first.

See also: why political science dismissed Trump and political theory predicted him; why political science dismissed Trump and political theory predicted him, revisited;

a civic approach to free speech

I argued in a recent post that libertarians, social democrats, American liberals, and most US Constitutional scholars share a sharp distinction between the state and the private sector–but this distinction does not reflect our actual experience of the social world.

One result is a certain way of thinking about freedoms of speech, the press, assembly, religion, and petition (the Five Freedoms of the First Amendment, which are also important rights in other democracies).

A typical first step is to identify which institutions are public or state bodies. They should be prevented from interfering with other people’s speech and assembly, and they should be constrained from expressing themselves in certain ways. For instance, the US government may not express support for any specific religion, although anyone else in the society may.

The next step is to safeguard the freedoms of non-public groups, including their freedom to discriminate and exclude. For instance, the Catholic Church is not required to ordain non-Catholics (or women) as priests. Such requirements would violate its freedom of assembly and religion.

Then we face two recurrent debates. One is whether various private associations (universities, web platforms) should act like states, even though perhaps they don’t have to under the Constitution. For instance, should a private university accord its students untrammeled freedom of speech? The other debate is whether hybrid institutions (state universities, political parties, public broadcasting services) are more state or private. Do they have First Amendment rights or must they safeguard others’ rights, or both?

The debate about the role of speech in our democracy thus centers on questions like comment-moderation, inviting or disinviting speakers, speech codes, hate speech–all of which have a legalistic flavor. The question is who has a right to say what, where.

If I actually had any influence, I would not seek to upset the apple cart of American constitutional thought. The categories that we have drawn (public/private, freedom/restriction) reflect some accumulated wisdom and offer some practical advantages. I would give a Burkean justification for how we employ the First Amendment: it is how we have learned to operate.

But the distinction between state and private sphere is at odds with the reality of how institutions work. They are almost all hybrids, partly public and partly private, exercising power but also allowing voice, including some and excluding others.

So what if we started instead with a population of people–individual human beings–who come together in a wide range of organizational forms to define, discuss, and address problems? I think these are the important points for them to consider in relation to freedom of speech:

  1. They need structured, reflective discussions that encompass a diversity of views and respond to good reasons or insights, not to power. They don’t need consensus, but they must continuously learn from others.
  2. Good discussions take institutional forms, from op-ed pages to seminars to town meetings. All institutions have rules, norms, resources, and incentives. Incentives are necessary because participation in a discussion has costs. It takes time and energy to discuss, and the conversation may cause discomfort. Individuals don’t have to participate. Successful institutions for communication or discussion find ways to lure people in. A classic example was the package of the local daily newspaper: comics and sports to encourage subscriptions, and a sober front page to direct your attention to serious matters. The demise of this business model is an important example of what we should worry about.
  3. Any good discussion is a common-pool resource. It requires voluntary contributions, it serves all who participate, but it is easy for individuals to ruin. There are principles for the management of fragile common-pool resources.
  4. On the list of principles you will not find a requirement to discuss all the rules and incentives all the time. On the contrary, groups must economize on disagreement. They can’t handle too much of it. And any discussion assumes a prior solution to a problem of collective action. People didn’t automatically want to show up and talk; they were drawn in. This means that discussions generally rely on founders, small groups of leaders, or past generations of participants. We don’t make our own discussions; we join them. The structure of the institution constrains the discussions that take place within it, but there is no such thing as an unstructured discussion.
  5. Given the fragility of institutions for discussion and the importance of building institutions that match various needs and interests, they must be plural. We need lots of overlapping but heterogeneous forums–face-to-face, online, big, intimate, ideologically coherent and ideologically diverse. Each one will set rules for what speech it allows, but the rules will also determine who participates, the costs and benefits of participation, the scale, and a range of other issues. No set of rules is ideal; it’s the whole ecosystem that matters.

None of this is original. It reflects well-developed lines of argument from the sociology of communication and other fields. But it is an alternative to the US discourse of free speech, which is all about rights and restrictions. It focuses instead on the design of multiple institutions for communication–their resources, boundaries, rules, and norms.

Sinclair and Bezos: media ownership and media bias

These two stories ran on the same page of the print New York Times on April 2: “Sinclair Videos Renew Debate Over Media Ownership” and “To Trump, It’s the ‘Amazon Washington Post.’ To Its Editor, That’s Baloney.”

Both articles are about possible bias in powerful, for-profit media companies. Donald Trump has opinions on each case. He thinks that Sinclair (which owns 193 local TV stations) is a “far superior” media company that is being smeared by liberals. But he suggests that the “Fake News Washington Post [is] being used as a lobbyist weapon against Congress to keep Politicians from looking into Amazon no-tax monopoly.”

I start with the opposite assumptions: Sinclair is a creepy would-be monopolist, while the Washington Post holds power accountable. I’m no fan of Amazon, but I assume that Jeff Bezos’ investments in the Post strengthen democracy by enabling the newspaper to do more investigative reporting. I see two threats to the First Amendment: Sinclair’s goal of owning more than 200 local TV stations, and a president–who sits atop a regulatory state–threatening the owner of a newspaper.

But imagine that you admired Trump rather than despising him (as I do). You might then reverse the polarity. The biggest threat might seem to be the billionaire with the national newspaper. You might be a little cynical when the Post‘s executive editor, Martin Baron, says, “There isn’t anybody here who is paid by Amazon … Not one penny.” Technically true, but Bezos, who makes his money from Amazon, bought the Post for $250 million, and, “Buoyed by [these] new resources, it has added more than 200 newsroom employees.”

I want to control corporate influence on politics, but when people point out that newspapers also influence politics, and they are corporations (or owned by corporations), I cry “First Amendment!” When Rupert Murdoch builds Fox News, I see a billionaire colonizing the public sphere, but when Bezos expands the Post, I am grateful to him.

It’s important to be principled, not arbitrary or reflexively partisan, in making such judgments.

The economics are complex. You can make money selling news that you don’t agree with, or lose money by selling views you like. (Generations of owners of The New Republic will testify to that.) There is evidence of media effects: news companies change public opinion. But consumers also choose which news to buy and thereby affect the production of news.

The sociology is complex. Sinclair Broadcast Group is a publicly traded company that maximizes returns for its shareholders. It is also an organization with a CEO and other leaders who have leverage over the shareholders. And it employs reporters, who can be understood as members of a profession that is committed to the public good. It would be naive to ignore the corporate structure, but cynical to ignore the professionals. An anonymous anchor interviewed by Maxwell Strachan said,

most of the people who are commenting on this have never even watched our local newscast. … They see that we’re a Sinclair station. They assume what they want to assume about it. But we produce good news here. Sinclair does not tell us what to cover, who to talk to, or what to say in terms of local coverage. Our local news, it doesn’t have bias. If people are looking for it, they won’t find it. So don’t call me a zombie. I do damn good work on a daily basis and anybody in my community would tell you that.

Whether he is right or not, his point that many critics have never watched a Sinclair Station’s local newscast applies to me.

Finally, the politics is complex. I have no doubt that Donald Trump is a terrible leader, but I choose to consume news that mostly reinforces that view, and I rarely delve deeply into the other side. We should make judgments and take a stand. Forming a judgment is not a form of bias. But we must recognize our fallibility.

Ultimately, you can’t render appropriate judgments without taking a closer look at both the products of these companies (Are their stories any good?) and the detailed ways in which they work. I presume that the Post has a strong firewall between its business operations and its newsroom, but that is an empirical assumption that can be tested. I find this kind of language in Sinclair’s employee handbook disturbing: Sinclair “may monitor, intercept, and review, without further notice, every employee’s activities using Company’s electronic resources and communications systems.” But I don’t know whether anything similar applies at the Post.

All of this plays out in a marketplace. We’d like journalists to have market power over their employers. But for newspaper reporters, the market is terrible:

According to a new Knight Foundation report, “In the decade since the last recession hit, newspapers have shed 26,300 newsroom employees — 46.1 percent of total employment. … In contrast, local TV news employment is up 4.9 percent in that same time frame, and most TV newsrooms are at their highest level of staffing ever.” Many stations are hiring former newspaper reporters and editors.

Despite the rising number of employees in broadcast journalism and the generally tight labor market, Sinclair has leverage over its employees, perhaps because there are just a few TV stations in any community. The anonymous Sinclair anchor says,

These jobs? they’re very hard to come by. And if I quit, I owe the company 40 percent of my salary, plus a percentage of the [redacted] years remaining on my contract, plus any bonuses that they’ve paid to me and any reimbursements that they’ve paid to me. And they’re going to take me to court for it. And in the time that I’m in court, I’m not employable.

See also: media literacy and the social discovery of realitydon’t confuse bias and judgment.

how information relates to power, according to C.V. Wedgewood

C.V. (Veronica) Wedgewood’s The Thirty Years War is almost a century old, but it remains an inexhaustible source of insights. TaNahisi Coates loves it, too: “Take this for whatever it’s worth but she writes better than any historian I’ve ever read. Like all of my favorite writers she paints in all colors. … This is just a thrilling book. Sometimes it’s too pretty, and the details are too on point, but the insights are so thorough and the narrative so gripping that it’s hard to turn away.”

Here’s an example. Wedgewood asks how dynastic politics–births and marriages–could have been so influential. The Hapsburg Empire, for example, was the greatest power in Europe and it formed because of royal weddings. “The dynasty was, with few exceptions, more important in European diplomacy than the nation. Royal marriages were the rivets of international policy and the personal will of the sovereign or the interests of the family its motive forces. For all practical purposes France and Spain are misleading terms for the dynasties of Bourbon and Hapsburg.”

(Wedgewood doesn’t mention the Ottomans, but they were also a family, not a people. The Ottoman Empire was proudly multinational, not Turkish, and it was defined by the fact that the Sultan was the lineal descendant of Osman I [1258-1326]. In Topkapi, marriages weren’t relevant, but it mattered which heir obtained the throne.)

How could the fates of millions be determined by who married whom in a few families?Wedgewood thinks the reason is information:

This is an interesting explanatory thesis. Perhaps it could be restated thus: Everyone has political interests. But in order to act on their interests, people need information and the ability to coordinate. Without information, the peasants and most of the middle class were rendered powerless ca. 1600. That left the great aristocrats to govern, and they could best understand and use their own relationships to shape the world. (They presumably had poor information about things like economics and demographics.) Their relationships were transparent to the masses–for example, everyone knew when the king got married–so the most likely point for popular involvement was in supporting or blocking a dynastic union.

This thesis also raises questions about our own time. Today, we have information by the gibibyte. What we lack is the ability to focus attention on the important stuff. It’s easier to grasp Donald Trump’s marital and extramarital relations than to follow how HHS is undermining Obamacare. One dominant man holds extraordinary power–and celebrity–in China, Russia, India, Turkey, and many other countries. It’s conceivable that the 21st century will look more like the 16th than the 20th in this respect.