A Modest Proposal for Supreme Court Reform

Mark Tushnet

The Supreme Court's argument calendars for January and February (quite light) provoke me to the following:

          The prospect of large-scale changes in the Supreme Court, never great, disappeared on November 3. Here’s a suggestion for a more modest reform that will get the Court to contribute its fair share to the coming push to tighten the belts of our institutions.

Cut the number of law clerks justices can hire from four to two (with one extra for the Chief Justice). In 1968-69 each justice had two clerks and the Court decided 122 cases. In 1972-73, when I clerked for Justice Thurgood Marshall, each had three clerks and the Court decided 164 cases. In 2018-19 each justice had four clerks and the Court decided 72 cases. That’s a drop from 55-60 cases per law clerk to 18 per law clerk. Maybe we were just a lot smarter than they are (but probably not twice as smart). Or maybe the opinions they’re turning out are twice as good as ours were.

Frankly, I doubt it. More likely, they’re just less productive. They spend more time going down rabbit holes doing unnecessary research. Or, maybe worse, drafting opinions explaining why their justice joins in Parts I, II (A), and III of  majority opinion but dissents from Parts II (B) and (C). (In one case Justice Kagan asked her law clerks to look up how every one of the fifty states used the words “tangible objects” in their statute books, just so she could include a footnote – well-written, to be sure – telling us the answer. Our law would have been none the poorer without the footnote.) 

If the Supreme Court were a private business and stockholders discovered that productivity had dropped in half while the staff remained the same size, they’d make management do something about it. And what management would do is trim the staff. Why should the Court be immune from efforts to cut the fat off an operation that’s gotten bloated as the justices themselves have chosen to cut down on the number of decisions they make? 

Any discussion of justices and their law clerks has to include the standard quotation from Justice Louis Brandeis: “The reason why the public thinks so much of the justices is that they are almost the only people in Washington who do their own work.” That hasn’t been true for decades, but cutting the number of law clerks would make it a bit more true.

The justices still do the most important thing: cast votes about who wins and loses. Sometimes they divvy up the work of drafting opinions, taking some for themselves, having law clerks draft others. The clerks try their best to capture what their justice thinks and to reproduce the justice’s writing style. Justices always “edit” the clerks’ drafts, though if the clerks have done their job well the editing should be light. 

How much of their own work the justices do depends in part on how long they’ve been on the Court. In a justice’s first years she or he does a lot of the drafting personally. As the justices get older and more experienced in managing their law clerks, drafting responsibility shifts toward the law clerks.For understandable reasons, no current or recent law clerks candidly describe how much drafting they actually did – and some actively misrepresent the facts. But you can see what practices were in the relatively recent past in the papers of several justices available at the Library of Congress and elsewhere. And “insider” scuttlebutt confirms that things haven’t changed on these matters. Because I was a law clerk (nearly fifty years ago) I personally don’t see anything wrong with the way the justices choose to run their offices – which is to say, how much input they have on the opinions that emerge from their chambers with the heading, “Justice X, concurring.” 

Still, decreasing the number of law clerks  will increase the amount of “their own work” that the justices do. And if Justice Brandeis was right, that will benefit the Court even if it burdens some of the justices.

 And, for those who thought that more substantial changes were desirable, consider the fact that as the burdens increase the job becomes less attractive, particularly as a justice grows older and tries to shift more work onto the law clerks’ shoulders. 

Cutting the number of law clerks in half might increase respect for the Court and increase turnover on the bench – not quite term limits, but getting closer to them.

Evaluating the Omnibus’s Anti-Poverty Measures

David Super

                Now that the President has deigned to sign Congress’s massive year-end opus, it seems appropriate to comment on how it responds to the impoverishing effects of the pandemic and recession. 

At the outset, I should note that one test commonly applied in popular media articles – stimulative effect – is largely inappropriate.  Job losses have been heavily concentrated in sectors that are now unsafe, such as travel and entertainment.  Bringing those jobs back is not primarily an economic challenge; it is a public health one.   To be sure, increasing aggregate demand by putting money into the hands of people likely to spend it should prevent additional job losses not necessitated by the public health crisis.  That effect, however, depends more on the size of the legislation than on its particular design. 

The economy will be slower than it needs to be in coming months because the strong anti-spending faction in the Republican Party insisted that the legislation stay well below the symbolic $1 trillion mark and because the (heavily overlapping) pro-tax-cut faction diverted a large share of the $900 billion to subsidies that are likely to stay on corporate balance sheets or to be distributed to high-income shareholders here and abroad, who will simply bank the gains. 

As relatively insignificant as the legislation’s design is for macro-economic purposes, however, from a micro-economic perspective its composition is quite important.  In particular, a large number of valuable assets are in imminent danger of destruction, resulting in waste and long-term hardship for those losing them.  These include homes, with evictions inevitably resulting in the loss of valuable personal property and the disruption of personal networks providing everything from job leads to spot child-care.  The undervaluation of these networks is a persistent failure in the design and appraisal of anti-poverty policy.  The continuation of the eviction moratorium, with language that will allow the Biden Administration to extend it further, will stave off many of these losses.  The inclusion of substantial, although woefully inadequate, rental assistance will allow many of these losses to be avoided altogether.

Another threatened long-term loss is education.  The same underfunded, over-populated schools that struggled most to serve their students in the best of times have had the least ability to help them during the crisis.  With the recession hammering state and local revenues from sales and income taxes, many of these schools have faced devastating additional cuts.  The legislation’s omission of significant state and local relief funding, beyond a very modest education aid package, squandered the chance of preventing these losses.  The pandemic’s legacy thus will include a further largely preventable widening of our education gap.

A third area of preventable long-term harm is from the coronavirus itself.  The expiration of enhanced unemployment benefits in late July, and the expiration of expanded eligibility for unemployment aid over the weekend, likely have forced many desperate low-income people to take jobs where they face a high risk of infection.  Some will become ill themselves and will bring the disease home to high-risk household members.  Lost in the debates over Republican claims that unemployment benefits discourage work – a claim with little empirical support in ordinary times – is the fact that we should want to discourage work in a pandemic where available jobs pose undue risks.  Employers can and should be pressed to compete for workers with safer working conditions.  Although the benefits’ duration is disturbingly short, and the halving of enhanced benefit levels from what the CARES Act provided in the Spring may induce some low-wage workers to endanger themselves and their families, rescuing these programs is a major accomplishment.  So is the rejection of Senate Majority Leader McConnell’s insistence that employers and others be immunized from litigation over unsafe working conditions.  On the other hand, the failure to extend workers’ right to take paid sick leave likely will force some workers exposed to the coronavirus to stay at work, endangering others.  Low-wage workers tend to work with other low-wage workers; forcing them to work while they may be sick likely will increase the disease’s spread within vulnerable communities. 

Beyond the macro- and micro-economic effects of the package, how it distributed the dollars it made available for low- and moderate-income people is important.  More so than in many relief packages, this legislation targeted aid on the most vulnerable.  Having set a grossly inadequate cap on the overall size of the package, Republicans tried to crowd out well-targeted unemployment benefits with larger scatter-shot economic impact checks.  This effort had only limited success, as evidenced by President Trump’s last-minute tantrum demanding much larger checks. 

Despite their mass appeal, rebate checks distribute a great deal of money on those in stable albeit not affluent circumstances.  This leaves much less for those in dire need who face the risk of losing housing, utility service, or food.  Many of the poorest of the poor were completely ineligible; many others had insufficient connection with the tax system to receive checks without making additional filings.  Herculean outreach efforts by many non-profits and state and local governments only helped a modest number of these people to make the necessary filings to receive checks.  Moreover, because the tax system depends on prior years’ data, it mismeasures current need:  income in pre-pandemic conditions is a deeply flawed indicator of unemployment and need during the crisis. 

The checks-versus-unemployment benefits debate also exposed a latent tension within the movement against income inequality.  One set of inequality critics, largely from the Left, focuses on the concentration of wealth at the top – the “one percent”, “the billionaires” – and redirecting some of that wealth to improve the lives of the middle class.  Many in this group are motivated by the corrosive political influence of such great wealth.  The other set focus on the growth of severe need at the bottom of the income spectrum, seeking to reduce the ranks of the poorest of the poor.  This group’s motivation is primarily humanitarian rather than political, and it draws adherents from a much wider range of the ideological spectrum. 

Although these two views of inequality are not irreconcilable in the abstract, in practice they represent competing claims on limited financial and political resources.  The acrimonious argument between Senator Joe Manchin – the chamber’s most conservative Democrat – and Independent Senator Bernie Sanders illustrated this split.  Sen. Manchin insisted that putting as large a share as possible of the available funds into unemployment assistance was the best way to target resources on those facing the greatest hardship; Sen. Sanders preferred giving a broader segment of the middle-class another taste of redistribution.   Like others, I found myself in the unaccustomed position of rooting Sen. Manchin on.  My sense is that many intellectual and political leaders on the Left take for granted the support of low-income people and their allies for the movement against income inequality and are puzzled when it is not forthcoming.  The movement’s future depends on its ability to overcome that myopia.

Nonetheless, the legislation is striking for several initiatives to reach the poorest of the poor.  Expanded eligibility for unemployment benefits focuses on many of the most marginalized workers, including those unable to find full-time work and those in the “gig economy.”  The enhanced unemployment benefit levels are most significant to the lowest-wage unemployed workers.  The 15% increase in the maximum benefit in the Supplemental Nutrition Assistance Program (SNAP) reaches the  poorest of the poor; by contrast, the Trump Administration’s contorted and unlawful interpretation of the SNAP increase in the Families First Coronavirus Relief Act completely excluded the poorest 37% of recipients and gave very little to other extremely low-income households.  The extension of the eviction moratorium and the allocation of additional funds to rental assistance also benefits many of the poorest of the poor.  On the other hand, Democrats failed to win an increase for the Temporary Assistance to Needy Families (TANF) block grant to provide for emergency grants to those in greatest distress whose needs other programs do not meet. 

Finally, the long battle over this legislation reflected the growing realization that state and local government finance is an important front in anti-poverty law.  This is not to say that state and local governments are reliably sympathetic to low-income people:  far from it.  But the federal government’s difficulty administering programs for low-income people, especially for the poorest of the poor, leave state and local governments the only hope for many.  (In a concession to those limitations, the omnibus legislation prevents earned income tax credits and child tax credits from dropping this year for laid-off workers:  workers can receive credits based on their 2019 earnings instead of those in 2020.)  State and local governments also fund non-profit community groups to help those that government refuses to aid directly.  And when state and local governments’ budgets deteriorate, the resulting cuts hit low-income people disproportionately hard.  Many state and local employees – 1.3 million of whom have lost their jobs since February – are themselves a few paychecks away from poverty.  The failure to include substantial new state and local aid in the omnibus legislation is a failure of anti-poverty law as well as of fiscal federalism more generally. 

The chaotic handling of expanded unemployment assistance, the failure to help state and local governments avoid mass lay-offs, and the Trump Administration’s forcing Congress to act a second time to get increased food assistance to the poorest of the poor all highlight the inadequacy of the ad hoc to economic crisis relief.  Although recessions do differ from one another, those differences are not so great that they require separate, individually tailored legislation each time.  A key argument against enacting permanent structures to address the effects of recessions is that Congress will inevitably come through when it is needed.  This is at least the third recession in a row when Congress’s inability to act caused great hardship for millions of unemployed workers and fell far short of preventing disastrous state and local government lay-offs.  The reforms in the major coronavirus relief laws should be enacted into permanent law with triggers that bring them on line automatically during a serious economic downturn. 

@DavidASuper1

Don’t über-blame the algorithms — lessons from Europe

Guest Blogger

From the Workshop on  “News and Information Disorder in the 2020 US Presidential Election.” 

Amélie P. Heldt 

One can hardly talk about disinformation and Europe without mentioning Brexit and the Cambridge Analytica scandal. However, the last four years have shown that blaming only technology for the information disorder is illusory. After Brexit, Europe feared disinformation campaigns would affect the French and the German elections in 2017. Until today there’s no evidence of such, either from foreign actors (e.g., so-called troll armies) or as a result of content curation systems on platforms. For instance, the 2017 German elections saw hardly any manipulation from abroad, and the influence of personalization by algorithms in search engine results was negligible, as Cornelius Puschmann wrote in “Beyond the Bubble: Assessing the Diversity of Political Search Results.” Similarly, false news sites had no significant influence on the 2017 French elections, researchers found. On the contrary, research shows that most of the issues are man-made and partly based on a failure of traditional media. Nevertheless, rumors concerning then-presidential candidate Emmanuel Macron perturbed the election campaign period and sparked off a regulatory reaction. 

Disinformation and the disingenuous discourse of victimhood

Guest Blogger

From the Workshop on  “News and Information Disorder in the 2020 US Presidential Election.” 

Ari Ezra Waldman

Donald Trump lies. He lies about many things. But most recently he has been lying about the 2020 election. He claims that if it weren’t for “fraud” associated with mail-in ballots in Pennsylvania, Georgia, Wisconsin, Michigan, and Arizona, he would have won the election. There is, of course, no evidence of systemic fraud. The only incidents of fraud identified in Pennsylvania were of a Trump voter trying to register another vote for Trump from his dead mother. And yet, surveys show that millions of Americans — almost all self-identified Republicans — believe that there was fraud, that mail-in ballots are fraudulent, and that but for this fraud, Donald Trump would have won reelection. Many elected Republican leaders have either actively supported these baseless claims or remained silent.

Scholars have probed the social, political, and technological aspects of misinformation, including its development, its spread, and its effect on democracy. I would like to discuss the Republican Party’s misinformation campaign about the 2020 election as a weapon of systemic racism, and its reflection of broader political and legal arguments in defense of white supremacy and traditional hierarchies of power.

Race, misinformation, and voter depression

Guest Blogger

From the Workshop on  “News and Information Disorder in the 2020 US Presidential Election.” 

Guy-Uriel E. Charles and Mandy Boltax

Since the 2016 election, scholars have focused a lot of attention on the intentional deployment of misinformation to influence electoral politics. The growing body of research on “information disorder” emphasizes our evolving social media landscape, which is an increasingly dominant environment for news that is driven by artificial intelligence, online behavior tracking, and data-analytics. Researchers have expressed concern about rising hyperpartisanship, a product of confirmation bias, motivated reasoning, and highly personalized social networks that minimize exposure to dissenting opinions and maximize positive feedback for certain beliefs. Other reports raise alarms about the shrinking market for facts in our increasingly dysfunctional political system, warning of foreign trolls, bots, and deepfakes sowing confusion and distrust, and domestic ideologues amplifying populist rhetoric that erodes institutional trust.

Academics have paid less attention, however, to misinformation directed to Black voters. There has been insufficient focus on how information disorder uniquely affects Black Americans, who have been disproportionately subjected to microtargeted content designed to demobilize, depress, and discourage their democratic participation. This lack of focus is unjustified particularly because disinformation targeting Black voters has been a distinctive electoral strategy of the Trump campaign.

The emerging science of content labeling: “Soft” interventions and hard public problems

Guest Blogger

From the Workshop on  “News and Information Disorder in the 2020 US Presidential Election.” 

John P. Wihbey

We are all content labelers — and potentially, the labeled — now. Indeed, we might think of 2020 as the dawn of information about information, the moral use of metadata in the market of speech.

Given this extraordinary turn toward labeling on social media, I want to focus here on a particular research agenda that explores a set of interrelated questions. They revolve around the tricky but fascinating problem of how to label information that may be problematic — incomplete, false, misleading, disputed, or otherwise in need of context. Answering these questions may be key to the organized, efficacious, and ethically justifiable governance of user-generated content on technology platforms, now and long into the future.

There are the narrower, tactical questions that everyone is asking right now about all of the content labeling we have just seen by Twitter, Facebook, and the like. Did any of those 2020 election labels “work”? How about the ones related to COVID-19? How might we define efficacy? How can we improve the user interface and user experience in this regard by tweaking, for example, the features, colors, and interaction design choices of the content labels?

Yet I want to reflect on deeper questions that are beginning to emerge around two areas: ethics and epistemology. These have come into focus as content labeling efforts have continually seemed haphazard, reactive, and often contradictory.

A Pocket Veto of the Stimulus Bill

Gerard N. Magliocca

A year that began with a presidential impeachment trial could end with another rare constitutional act (at least in recent decades): a presidential pocket veto.

The pandemic stimulus bill passed both Houses of Congress on Monday, December 21st. The Constitution gives the President ten days (Sundays excepted) to decide whether to sign or veto a bill. If he chooses not to sign or to veto, then normally the bill becomes law. If the Congress adjourns during the ten day period before the President acts, though, then the bill dies.

Ten days excluding Sundays from December 21 brings us to January 1st, right before the new Congress comes into session. This means that the lame-duck Congress can stay in session and block a pocket veto, though I don't know if Congress has ever stayed in session into a third January since the starting date was moved to January 3rd by the 20th Amendment. If the President takes the full ten days to act, though, foes of the bill could try to use stalling tactics to delay an override vote until one or both Houses must adjourn. sine die. 

One last thought--the pocket veto rule means that anything Congress might pass after Christmas must receive the President's signature to become law. He need not issue an actual veto to block the bill. 

UPDATE: I was incorrect in one respect. The stimulus bill is not yet enrolled. Thus, the 10 day clock is not yet running. As a result, the President can pocket veto the bill. I decided to leave the original post as is.

The 2020 election integrity partnership

Guest Blogger

From the Workshop on  “News and Information Disorder in the 2020 US Presidential Election.” 

Jevin West

“They gave Trump voters sharpies and now their votes are being invalidated! WTF!”

An unverified user posted this tweet shortly after the Arizona polls closed. Within a few hours, the tweet and others like it went viral. The official (verified) Pima County account tweeted a response to Sharpiegate, stating that felt-tipped pens could be used to vote; they would not invalidate a ballot. That was not enough to stop the surge. Hundreds of thousands of tweets followed, pushing the narrative of voter fraud.

Sharpiegate teaches us several lessons. Domestic disinformation was far more potent and prevalent in 2020 than inauthentic, foreign actors. Political operatives and social influencers in the U.S. didn’t need to create stories of voter fraud. They just needed to amplify existing ones, like this one in Arizona by an authentic, domestic user.

Sharpiegate is one of hundreds of conspiracy theories that my colleagues and I have tracked over the last year. On Dec. 3, 2019 — exactly one year ago today — we launched the Center for an Informed Public (CIP) at the University of Washington with support from the Knight Foundation. Since then, we have spent day and night monitoring misinformation surrounding COVID-19, vaccine hesitancy, the West Coast fires, the social justice movement, and now the U.S. election. Like other researchers, we have seen the rise of Plandemic, 5G, and Sharpiegate. We have seen the actors and tactics from one conspiracy theory emerge within the next. We have seen social media platforms experiment with banners, tags, and takedowns. And we have seen policy makers begin to take notice, writing laws limiting the use of synthetic media and other forms of deceptive technology. The year 2020 has delivered a decade’s worth of research material.

What the COVID-19 Relief Package Tells us about Congress

David Super

     As the 116^th Congress staggered toward the oblivion it so richly deserves, it offered up a gigantic year-end legislative package that funds the government for the remaining nine-plus months of the fiscal year, provides some relief to those hardest-hit by the pandemic and recession, and attends to a host of other bits of business, large and small.  Both the behemoth itself and the process bringing it about represent potentially significant departures from past patterns that may portend changes in the way policy is made going forward.  This post offers some observations about each.

Regulating AI: The question now is no longer whether, but how

Guest Blogger

From the Workshop on  “News and Information Disorder in the 2020 US Presidential Election.” 

Olivier Sylvain

The distribution and availability of networked information devices, applications, and services define the ways in which consumers share with friends, transact business, learn, and create. We did not need the pandemic to reveal this. I and others have for decades been writing about the ways in which a robust, resilient, and equitably distributed internet infrastructure is vital to the successful operation of democracy and free markets.

Just as salient for policymakers today is whether, now that the availability of internet structure is widely understood to be essential, policymakers can or should do anything more to ensure that the applications and services that such systems now make available abide by longstanding democratic norms and consumer protections. This question remains unresolved largely because, for the past quarter century, the preponderance of technologists, scholars, and policymakers have been skeptical about government regulators’ competence and capacity to promulgate timely or effective rules. The prevailing view presumes that such interventions are likelier to impede technological innovation than achieve any well-meaning objective. Thus, today, policymakers are reticent to do anything that risks slowing invention and entrepreneurship in the market for network information services. This is the view that Congress and the courts have enshrined in a handful of regulatory regimes about which I have written, including the FCC’s regulation of broadband network management, the court-made doctrine under Section 230 of the 1996 amendments to the Communications Act, and the public regulation of social automated decision-making generally.

Disinformation gets physical: The internet of things as an emerging terrain

Guest Blogger

From the Workshop on  “News and Information Disorder in the 2020 US Presidential Election.” 

Laura DeNardis

I can live, and live abundantly, as a digital human being who has never been on Facebook or Instagram. But I can’t live, work, or socialize without the underlying systems of technical architecture and governance that keep the internet operational and connect the cyber-physical world around me. The institutions that operate this infrastructure include cloud computing providers, internet registrars and registries, transaction and financial intermediaries, network operators, hosting companies, DNS resolutions providers, certificate authorities, and internet of things (IOT) system operators, to name just a few categories. They are not immediately visible to end users in the same way content and applications are visible. Yet they are the internet’s most powerful control points.

Understanding both vexing disinformation problems and powerful solutions to these problems requires a deeper dive beneath social media into underlying infrastructure points of control and the companies that operate at these various layers. As I have written about in the past (e.g., here, here, and here), the internet’s core infrastructure provides choke points — for better or worse — at which content and transactions can be blocked, altered, or co-opted. The Indian government’s approach to shutting down the internet or China’s approach to censorship and control exemplify how internet architecture is a now proxy for geopolitical power, as well as every manner of content control from DNS-based copyright enforcement to cloud computing-based filtering of child pornography. During the rise of COVID-19 disinformation, web hosting company Squarespace took down the website of the so-called America’s Frontline Doctors organization for violating its terms of service by making spurious claims such as not needing masks to battle COVID-19.

As more things than people are now connected to the internet, what is less obvious is how the internet diffusing into the material world all around us — the so-called internet of things — connects to disinformation. Far more “things” than people connect to the internet. These objects — whether Wi-Fi connected medical devices, consumer IOT, industrial cyber-physical systems, or smart city infrastructure — exist simultaneously in both the material and cyberworld.

In my new book, “The Internet in Everything: Freedom and Security in a World with No Off Switch,” I note that “all of the policy issues in two-dimensional digital space have leapt into three-dimensional real-world space and have added new concerns around physical safety and everyday human activity.” The internet of things has not yet been drawn into policy debates about disinformation, and this should change. The IOT is an emerging terrain of disinformation, and possibly one that is more consequential than social media influence campaigns, computational propaganda, video deepfakes, or any manipulation involving human content and communications. The following are three broad categories of disinformation arising via the co-option of IOT infrastructure:

The evolution of computational propaganda: Bots, influencers, and platform responsibility

Guest Blogger

From the Workshop on  “News and Information Disorder in the 2020 US Presidential Election.”

Samuel C. Woolley

When my colleagues and I began studying “computational propaganda” at the University of Washington in the fall of 2013, we were primarily concerned with the political use of social media bots. We’d seen evidence during the Arab Spring that political groups such as the Syrian Electronic Army were using automated Twitter and Facebook profiles to artificially amplify support for embattled regimes while also suppressing the digital communication of opposition. Research from computer and network scientists demonstrated that bot-driven astroturfing was also happening in western democracies, with early examples occurring during the 2010 U.S. midterms.

We argued then that social media firms needed to do something about their political bot problem. More broadly, they needed to confront inorganic manipulation campaigns — including those that used sock puppets and tools — in order to prevent these informational spaces from being co-opted for control — for disinformation, influence operations, and politically-motivated harassment. What has changed since then? How is computational propaganda different in 2020? What have platforms done to deal with this issue? How have opinions about their responsibility shifted?

What can and should platforms be responsible for?

Guest Blogger

From the Workshop on  “News and Information Disorder in the 2020 US Presidential Election.”

Daniel Kreiss and Bridget Barrett

In the interdisciplinary field of misinformation and disinformation studies, there has been no consistent answer to the question of ‘what can and should platforms be responsible for in the context of democratic decay’ (i.e., a context in which democratic institutions, structures, norms, and governance are eroding)? In this post, we take up this question in light of several relevant literatures on democracies in transition towards more authoritarian or ‘hybrid’ regimes, or more broadly, literatures examining the deterioration of core democratic processes, institutions, and governance mechanisms. 

Our review of this literature reveals that we lack good theoretical and empirical understandings of media systems in relation to democratic decay, especially the roles played by platforms. To address this, we conceptually outline several indirect effects of platforms on democratic decay, focusing on their roles in shaping public opinion and political institutions. Our goal is to bring two academic literatures together: 1) work on democratic decay that often fails to consider media and platforms, and 2) work on platforms that often focuses narrowly on public opinion and attitudes, overlooking institutional democratic processes.

A model for intuitive internet governance

Guest Blogger

From the Workshop on  “News and Information Disorder in the 2020 US Presidential Election.”

Kate Klonick

The question presented to this panel as a point of discussion was, “What is the platform’s role and responsibility for information disorder during the 2020 presidential election season?” That’s a fairly specific and narrow question — and arguably one tainted with the bias of hindsight. But it does serve as an essential case study of how to answer the general inquiry that is closing this conference: “Which internet architecture or platform governance system would best contribute to resolving information disorder in online and social media environments?”

This essay aims to introduce a model for understanding when platforms have a responsibility to create governance systems that adhere to core democratic values, such as transparency, accountability, and participation.

There is no one answer to the question of “best” internet architecture or platform governance system, just as there is no “best” architecture for running a small business or structuring a constitutional court. The precise architecture of a governance system will depend on (1) what you are trying to govern and (2) at what scale. For instance, the governance structure around municipal garbage removal has very different implications than the governance structure around the board of elections. Garbage removal is a service provided by a government agency, not an entitlement or right. In contrast, elections and voting are rights protected by an entity like a board of elections. Government systems that are transparent and participatory are more critical where rights are concerned, while outcome might be more important where the subject matter of governance is around a service. In either case, the government architecture will likely change not just based on subject matter, but scale. The architecture of governance of a small seaside resort town in Massachusetts bears little resemblance to that of Kansas City, never mind the European Union.

The answers to the above questions will then define expectations around the role of the platform and how robust the platform governance system should be in its commitment to the values of democratic governance. This graph visually represents this idea:

The road to hell is paved with good algorithms: The case for deactivating recommendation algorithms in the political sphere

Guest Blogger

From the Workshop on  “News and Information Disorder in the 2020 US Presidential Election.”

Jonas Kaiser

Algorithms, and recommendation algorithms in particular, are deeply ingrained in our networked public sphere. Facebook recommends us new friends, pages to like or groups to join, Google websites that fit our search interests, Twitter people to follow or topics to check out, Amazon products to buy, Spotify music to listen to, and YouTube videos to watch.

According to Alexa, the most visited websites in the U.S. are Google, YouTube, Amazon, Yahoo, and Facebook. Recommendation algorithms are an integral part of each of those websites. As Eli Pariser, Safiya Umoja Noble, Frank Pasquale, and many others have argued, there are myriad reasons to be concerned about algorithms’ integral role in our daily  lives. These include the creation of homogenous communities without our knowledge; reproduction of racism and misogyny; and the concealment of algorithmic decisions to begin with. Against this background, as well as my own research on the far-right, disinformation, and algorithms, I argue that social media companies need to deactivate their recommendation algorithms in the political sphere. I structure this demand around our knowledge of how machine learning works, the pitfalls of automated content curation, how corporate goals can run counter to the public good, as well as findings from my own research.

When thinking about recommendation algorithms, we need to think about machine learning, and statistical probability models. After all, this is what algorithms are. Yet, as British statistician George Box famously said: “All models are wrong, but some are useful.” As Momin Malik highlights convincingly, these models are approximations, and there are numerous ways in which they can fail or have shortcomings. On a more applied note, Harini Suresh and John Guttag identified five biases that compromise algorithms: historical bias, representation bias, measurement bias, aggregation bias, and evaluation bias. These biases can lead to problematic outcomes that might reproduce issues such as racism, misogyny, and more. In short: algorithms, no matter how good, will always have limitations.

Coordination: A prerequisite for an effective fight against misinformation

Guest Blogger

 From the Workshop on  “News and Information Disorder in the 2020 US Presidential Election.”

 Valerie Belair-Gagnon, Oscar Westlund, and Bente Kalsnes

At the beginning of the 2020 U.S. election, Twitter marked false or misleading information. It then changed its strategic orientation by hiding fewer false or misleading posts to contextualize fact-checks, such as by linking to the source of the information. This example confirms how platform companies have become arbiters of truth, while news organizations and fact-checking companies are seeking to regain gatekeeping power from Big Tech and Silicon Valley.

This particular arbitration of “what is truth” shows a stark contrast with what journalists have sought to achieve. Journalists have been described as civic gatekeepers, meaning “civic and moral roles of journalistic institutions, and their enactment of cultural codes that give shape to, and help to protect, a society’s normative values.” Meanwhile, platform companies have taken the role of gatekeepers of democracy, a role traditionally taken by formal political actors through formal political channels and engaging in opinion power, as Helberger argues. These two forms of gatekeeping shows a detachment between media, tech, and platform companies, the latter having demonstrably been developing policies on the fly. Meanwhile, governments are increasingly trying to regulate and limit platform power and, as Meese argues, they may fail to address the interconnectedness of platforms and news publishers, and ultimately care for public interest.

These power dynamics between fact-checkers, journalists, and platform companies prompt a set of questions in the fight against misinformation. Who are the winners and losers? Who does what for what gain? How can people find truthful information? In other words, what does the digital labor of fighting against misinformation tell us about the technologically driven practices and ways in which actors are seeking to gain legitimacy with their audiences? And “what are the limits of what fact-checking can accomplish without greater support from platform companies for the researchers, journalists, and fact-checkers seeking to understand and limit the spread of harmful misinformation?”

Facebook’s responsibility

Guest Blogger

From the Workshop on  “News and Information Disorder in the 2020 US Presidential Election.” 

Ethan Porter

People do not come to believe in misinformation out of nowhere. First, they must be exposed to the misinformation; then, they must not be exposed to a correction of the misinformation. Viewed in this light, belief in misinformation is a later step in a long process that implicates not only human psychology, but the architecture of online platforms. These platforms, too, do not come from nothing. Their design reflects human decision-making — decisions that can have profound effects on whether or not people believe misinformation.

Decisions that Facebook has made about fact-checking illustrate how particular decisions about platform design can facilitate belief in misinformation. Some of the company’s less defensible choices in this area are well-known. Company leaders, for example, have reportedly prevented misinformation disseminated by political figures from being corrected. The company is aware that its labeling of various posts by Trump as false has proven ineffective, but it has done little to curb misinformation spread by Trump and his allies. In the words of one top executive, it is not the company’s role to “intervene when politicians speak.”

Good-enough interventions

Guest Blogger

From the Workshop on  “News and Information Disorder in the 2020 US Presidential Election.” 

Lisa K. Fazio

We live in a culture where the truth has been devalued — where politicians who repeat falsehoods win elections, where lies spread further and faster than the truth, and where misinformation, conspiracy theories, and junk science run rampant on YouTube, Facebook, Twitter, and many other online spaces.

This is a huge problem. And an overwhelming one. It’s often easier to simply document the problem, wring our hands about how difficult it is to solve, and do nothing. In fact, over the past few years, as researchers and the media have done a lot to document the problem, social media companies and politicians have done very little to solve it.

Twitter and Facebook’s actions during the recent U.S. election of more aggressively labeling false claims, promoting quality news sources, and preemptively debunking voter misinformation are great first steps. But many of the interventions are being removed after the election, and there’s a complete lack of transparency. Overall, social media companies are still allowing, and in some cases encouraging, the spread of false information on their platforms.

The good news is that simple, easy-to-implement solutions can help solve this problem. The issue is that none of them, on their own, will fix it. Instead of holding out for the perfect solution, we need to start implementing numerous small changes to our political system, social media platforms, and our own behavior. We need to start implementing “good-enough interventions.”

Social influence campaigns in the cyber information environment

Guest Blogger

From the Workshop on  “News and Information Disorder in the 2020 US Presidential Election.” 

Kathleen M. Carley

Beliefs, opinions, and attitudes are shaped as people engage with others. Today, that is often through social media. A key feature of social media is that all information is digital, and all information is shared through devices by agents who need not be human. This has created a Wild West for information where it is as easy to create and share false information as true, where nonhuman agents are often better equipped than humans to share information, where technology arms races continually alter the landscape as to what is doable, and where human understanding of the phenomena and policy are lagging.

Lone wolves and large propaganda machines both engage with the public to disrupt civil discourse, sow discord, and spread disinformation. Bots, cyborgs, trolls, sock puppets, deepfakes, and memes are just a few of the technologies used in social engineering aimed at undermining civil society and supporting adversarial or business agendas. How can social discourse without undue influence persist in such an environment? What types of tools, theories, and policies are needed to support such open discourse?

News organizations as fact-checkers: Any potential issue?

Guest Blogger

From the workshop on  “News and Information Disorder in the 2020 US Presidential Election,” 

Jisu Kim and Soojong Kim 

During the presidential election, major news organizations, including USA Today, The New York Times, The Associated Press, and The Washington Post, have verified with fact-checking organizations what politicians said in their speeches and presidential debates. However, unlike fact-checking organizations, which are not strongly connected with a specific group of audiences, several news organizations have been preferred and trusted by different audiences. For example, CNN and Fox News, two major political news sources, are differentially trusted by these audiences. According to Pew, more than 60% of Democrats trust CNN, while more than 60% of Republicans trust Fox News. Only 20% of Republicans got their political and election news from CNN; 60% of Republicans got their political and election news from Fox News in 2019. 

Considering the public’s different preference levels toward individual media organizations based on their political predisposition, how does the public differentially perceive and share news articles that correct misinformation about politicians? If the public perceives a relationship between a news organization and a politician as favorable, are their evaluations regarding the credibility of the news reports that fact-check or correct misinformation about one of the politicians different, depending on the news organization?

That Op-Ed About Jill Biden Is Awful. Northwestern’s Response Might Be Worse.

Andrew Koppelman

Steve Lubet and I have a new piece in the Chronicle of Higher Education on L'Affaire Epstein. It is, unhappily, paywalled, but here is the opening:

Misinformation research, four years later

Guest Blogger

From the Workshop on  “News and Information Disorder in the 2020 US Presidential Election,” 

Andy Guess

The shock of Donald Trump’s unprecedented election corresponded to a flurry of activity among researchers, civil society groups, and foundations to understand what they missed about social media’s role in fostering a degraded information ecosystem. The result was a surge of research about various forms of misinformation, which went hand in hand with increased scrutiny of social platforms.

Since that time, an extraordinary amount of effort went into preparations to avoid a repeat of the surprises of 2016. Meanwhile, the pandemic provided the additional challenge of combating rampant misinformation about the virus during a time of uncertainty around a constantly evolving global health threat. A somewhat surprising result of all this has been a willingness by platforms to employ increasingly aggressive tactics designed to reduce the spread of misleading content, such as claims of voter fraud and vaccine skepticism. At the same time, a number of anticipated threats — major foreign influence operations, deepfakes — did not seem to emerge.

Now that the election has passed and efforts to subvert the outcome have failed, where does all this leave misinformation researchers? Before offering some possible answers, I will begin with a quick overview of how I think scholars responded to newly salient questions about the role of online misinformation in political behavior over the past four years.

Towards a diachronic understanding of the harm potential of information disorder — reflections on Election 2020

Guest Blogger

From the Workshop on  “News and Information Disorder in the 2020 US Presidential Election,” 

Sam Gill

The study of information disorder is a study in epistemic anxiety. The anxiety that false information may be harmful to our democracy is matched by an anxiety that we have little insight into prevalence and effects of exposure to false information. This, in turn, is matched by an anxiety that digital information platforms are simply funhouse mirrors, distorting and enlarging analog social pathologies. The liturgy then completes its recursive turn in anxiety that we know truly nothing without greater access to platform data, concluding in a supplication to the platform companies. (Or some higher power — for example, a European Union consultative process).

Amen.

Yet what the 2020 election has accentuated is an opportunity for progress, both in assessing the moral harm from information disorder and in measuring it. The key is in identifying the species of harm that are of the highest moral urgency as the starting point for observation. In particular, the material experience of the election suggests that digitally disseminated, accelerated and amplified false information should be understood “diachronically,” to borrow a term from linguistics. That is, the time, place, and manner of false information all matter, as do its provenance and understanding of those most affected. History, in a wide sense, is meaningful for grasping both the mechanics of false information and the magnitude of moral harm.

Yale Information Society Project Conference on News and Information Disorder in the 2020 Presidential Election

JB

On December 4th, the Yale Information Society Project held a workshop on  “News and Information Disorder in the 2020 US Presidential Election,” organized by Yale ISP Resident Fellow Jisu Kim. ("Information disorder" is a catchall phrase to describe the various forms of misinformation, disinformation, and mal-information.)

Over the course of the next week we will be posting short papers based on the presentations given at the conference on Balkinization. The papers will also be published at the Knight Foundation.

Here are YouTube videos of the sessions.

Session One : The Creation and Spread of False Information, and Political Discussion

Session Two: How to Fight Information Disorder I

Session Three: How to Fight Information Disorder II

Session Four: Platforms' Roles in Resolving Information Disorder

Session Five: Algorithms, Bots, Internet Architecture and Governance

Session Six: Law and Policy to Resolve Information Disorder

Walmart Leads the Way

Ian Ayres

 Ian Ayres, Zachary Shelley, and Fredrick E. Vars

              In the lead up to the election, Walmart removed guns and ammunition from sales floors of its U.S. stores that sell firearms, in attempt to prevent theft of firearms if stores are looted amid civil unrest.  And then, when its assessment of the likelihood of civil unrest ebbed, it nimbly changed course and returned these items again to its shelves.

              These sensible moves are far from the first time Walmart has shown leadership in trying to assure that it was responsibly marketing products that can be violently misused.  Last year, it stopped selling ammunition that can be used semiautomatic rifles and the year before it raised the minimum age to purchase guns or ammunition to 21.

But can a single gun retailer make a dent in gun violence?  Our new research suggests that it can.

             In 1994, Walmart stopped selling handguns at all of its locations in every state except for Alaska. Then in 2006, Walmart stopped selling firearms altogether in more than half of its stores.  But the company partially reverse course in 2011 and began increasing the number of stores selling rifles and shotguns.  We tested the impact of these policy changes on suicide and homicide.

              There are reasons to be skeptical that Walmart’s policy changes would have a significant impact.  The United States is awash with gun dealers.  There are more than 62,000 federally licensed dealers, more than the number of grocery stores or pharmacies.  But not all dealers are created equal.  In any given month, more than two-thirds of Americans visit a Walmart store. Even with its self-imposed restrictions, Walmart remains the nation’s largest gun dealer. 

              Another reason for skepticism about the possible impact of Walmart’s decisions is that many gun suicides involve someone else’s firearm or firearms purchased years earlier.  Still, a substantial number of suicides do involve recently acquired firearms.  And making gun acquisition even marginally more difficult could reduce gun ownership over time.  Having a firearm in one’s home substantially increases the risk of suicide.

              In fact, in a forthcoming statistical analysis we find that Walmart’s 1994 decision to stop selling handguns reduced firearm suicides without increasing non-firearm suicide.  From 1994 to 2005, controlling for a variety of legal, social and demographic variables, counties with Walmart stores experienced a 3.3 to 7.5% reduction in the gun suicide rate (without an increase in non-gun suicides).  Our estimates suggest that Walmart’s decision to stop selling handguns has saved between 425–998 lives every year.  Between 1994 and 2005, this represents more than 5,000 lives saved.  On the other hand, Walmart’s 2006 and 2011 decisions to discontinue then resume the sale of rifles and shotguns in many of its stores did not significantly impact suicide.  The greater effect of the 1994 no-handgun policy makes sense.  Seventy-five percent of firearm suicides involve handguns. 

              The Walmart example shows that the decisions of private businesses can significantly reduce gun violence.  This is not to say that public policy has no role.  To the contrary, many public policy interventions have been demonstrated to reduce gun violence, especially gun suicide.  Gun sales have surged during the COVID-19 pandemic, raising concerns about increased risk of suicide and domestic violence homicide. But public policies have mitigated the surge in some states.  For example, states like New York where gun shops have been closed as non-essential businesses and states with licensing requirements have experienced a reduction in firearm sales.              

Walmart’s recent decision to restrict firearm displays should be applauded because it reduces the chance that firearms from its stores will be looted and fall into the hands of individuals who would misuse them.  But our estimates underscore the possibility that corporations can play a broader role in mitigating the country’s gun suicide crisis.  Walmart has led the way over the years with a series of self-imposed firearm sales restrictions.  Its decision in 1994 to stop selling handguns likely has saved thousands of lives.  Customers and employees at other substantial retailers, such as Bass Pro Shops and Cabela’s (which still sell handguns) and Dick’s Sporting Goods (which has eliminated firearm sales at some of its stores), would do well to take note.  It is time for more corporate executives to show leadership on reducing gun violence.