Comparisons between today and 1930s Nazi Germany are legion. Hardly a day passes without someone comparing Trump’s praise of Twitter as a way to reach the people directly, to Nazi Propaganda Minister Joseph Goebbels’ purportedly similar praise of radio. In 2017, Daniel Ziblatt drew on his political science work about conservatives in the Weimar Republic (and their use of media) to coauthor a popular book with Steven Levitsky about How Democracies Die.1 That same year, Timothy Snyder wrote a pamphlet book with twenty rules for how to survive fascism, drawing from his work on the 1930s and World War II.2
This does not mean that today is destined to be a rerun of the interwar period. But the resonances suggest historical patterns. These patterns can make us more critical about assertions of radical novelty in the present. If we fall into the trap of believing the novelty hype, we miss multiple important points. First, we might forget the path dependency of the current Internet.3 Second, we might misdiagnose contemporary issues with social media platforms by thinking about them too narrowly as content problems, rather than within a broader context of international relations, economics, and society. Third, we might focus on day-to-day minutiae rather than underlying structures. Fourth, we might think short-term rather than long-term about the unintended consequences of regulation. Finally, we might inadvertently project nostalgia onto the past as a “Golden Age” that it never was.
Some aspects of the Internet are unprecedented: the scale of its reach, the microtargeting, the granular level of surveillance, and the global preeminence of US-based platforms. But many patterns look surprisingly familiar – for instance: oligopolistic companies, political influence, and short-term thinking that focuses on media above and beyond broader societal problems. This chapter will explore five patterns from history that can help us to understand the present.
I developed the framework in this chapter for my testimony before the International Grand Committee on Big Data, Privacy, and Democracy in Ottawa, Canada in May 2019. This committee was formed in fall 2018, when the British Digital Culture Media and Sport (DCMS) committee gathered together twenty-four representatives from nine countries for a hearing. The DCMS committee had been investigating the role of Facebook and social media in the Brexit referendum. In a highly unusual move, the British committee had travelled to Washington, D.C. to question representatives from social media companies. The committee had subpoenaed Mark Zuckerberg to appear before them in the United Kingdom. Zuckerberg declined. In response, Britain teamed with Canada’s Standing Committee on Access to Information, Privacy and Ethics (ETHI) and representatives from nine countries in total for hearings in London in November 2018.4 The second committee meeting in Ottawa included representatives from Canada and ten other countries, ranging from St. Lucia to Mexico, to Estonia. Again, Mark Zuckerberg and Sheryl Sandberg were subpoenaed and they did not appear.5 When invited to testify before the committee, I worked on a framework that would provide a usable history for policymakers, but not one that simplified for the sake of political point-scoring. It is all too tempting to create a highlights reel from the past; it is far more productive to examine the history and bring that as evidence to the table.
Historian Sam Haselby has suggested a key distinction between history and the past:
Think of history as the depth and breadth of human experience, as what actually happened. History makes the world, or a place and people, what it is, or what they are. In contrast, think of the past as those bits and pieces of history that a society selects in order to sanction itself, to affirm its forms of government, its institutions and dominant morals.6
This chapter uses history rather than the past to discuss five patterns in the relationship between media and democracy. The history does not provide simple lessons that can be applied universally regardless of context. Instead, the history of media and democracy is messy and often counterintuitive. It often does not offer politically convenient answers. What history can give us, is a long-term perspective, a way to ask broader questions, and another analytical approach to the current moment.
Five Historical Patterns
1. Disinformation is an international relations problem.
Information warfare may seem new. In fact, it is a long-standing feature of the international system. Countries feeling encircled or internationally weak may use communications to project international prowess. This was as true for Germany in the past as it is for Russia today. We are returning to a world of geopolitical jockeying over news. If the causes of information warfare are geopolitical, so are many of the solutions. These must address the underlying foreign policy reasons for why states engage in information warfare. To address this, we need to understand when and why states use information warfare to achieve geopolitical goals.
Germans, for example, did not always care about international news. In the 1860s and 1870s, Germany was just unifying into a nation-state. Chancellor Otto von Bismarck cared about international relations. But he also cared about achieving German unification and then maintaining Germany’s status within Europe. Bismarck tried to influence journalists, particularly in London, Paris, and Berlin. He also intervened to ensure that Germany had its own semi-official news agency. But Bismarck did not mind that the global news supply system developed in such a way that British and French firms collected and disseminated most of the news from outside Europe, even for the German news agency.
Only from the 1890s, did German politicians and business owners start to care about and disagree with this system. They believed that it enclosed Germany at a time when the country wanted to become an imperial and global power. The news supply system had not become less effective from a media perspective. It had become so from a political perspective. Germans turned to information to push this agenda: many Germans were convinced that they had lost the world war of words and now needed to send news around the globe to counter Allied propaganda.7
For a historian, it is strange to see Americans so surprised that information falls under foreign policy. There is a long, often forgotten history of “active measures” or disinformation.8 “Psychological warfare” was a key concept for the CIA during the Cold War and the Department of Defense during the Vietnam War.9 After the Vietnam War, the Carter and Reagan administrations both incorporated information into their national security strategies. By 2000, these strategies for active engagement abroad were known in the Department of Defense’s Joint Staff Officers’ Guide under the acronym of DIME: diplomatic, informational, military, and economic power.10
This historical perspective makes recent Russian efforts seem less of an anomaly. If information has long formed part of international relations, we should not be surprised to see Iran, Saudi Arabia, and other states using social media to fight perspectives they dislike.
2. We must pay attention to physical infrastructure.
It seems so easy to access information on smartphones and wireless devices, that we forget the very physical infrastructure underpinning our current system. That current system also perpetuates inequalities in communication stretching back at least to submarine cables and steamships carrying the post in the mid-nineteenth century.
The first submarine cable was laid between the United Kingdom and France in 1851. After two unsuccessful attempts, a transatlantic cable was completed in 1866. In the interwar period, Austrian writer Stefan Zweig would pick that event as one of his Sternstunden der Menschheit (Decisive Moments in History).11 Cables spread rapidly around the world. But they followed specific patterns. Instead of connecting previously unconnected places, they created denser networks where networks already existed. Cables quickly connected British imperial territories to London. The Atlantic soon housed the most cables. The major company laying cables was a conglomerate, the Eastern and Associated Telegraph Companies, headquartered in London, but with Anglo-American financial backing.12
The company focused on places that seemed profitable. Unsurprisingly, these were places with trade connections. Cable entrepreneurs laid cables where business already existed. In one instance, the managing director of the biggest multinational cable company, James Anderson, argued against a proposed cable from the Cape of Good Hope to Australia, via Mauritius. He said the Eastern and Associated Telegraph Companies simply did not lay cables where there was “not even a sandbank on which to catch fish.”13 Market orientation shut out connections where massive profits could not be made.
Cable entrepreneurs differed from current social media platforms in one key way: men like Anderson thought that telegraphy was a communications medium for elites and that most people simply would not pay for international telegrams. Telegrams were highly priced and only about ninety businesses made regular use of transatlantic cables in the first few decades of their existence (alongside governments and the press). Cable entrepreneurs subscribed to the paradigm of high-cost, low-volume, which differs from today’s social media unicorns who seek rapid growth and billions of users above all else. But those cable entrepreneurs created infrastructure systems that have influenced communications networks until today.
These apparently global communications infrastructures had imperial roots. Africa in particular seemed less important for telegraph companies, because there would be fewer high-paying clients than in white dominions. Britain’s “All-Red Route” around the world was completed in 1902 and enabled the British to send cables around the world while only touching on imperial soil. Of the entire African continent, the cables only landed in South Africa. Other cables spread up the coasts of Africa but with far less density than across the Atlantic. Racist beliefs about African colonial subjects’ inability to communicate dovetailed with imperial communications governance.
Submarine cables set precedents for later communications networks in the twentieth century, like telephone cables and fiber-optic internet cables.14 Cables were generally laid on ocean beds that had already been explored, as this saved money. This also followed the pattern of laying cables where proven markets for communication already existed. Fiber-optic internet cable networks resembled submarine cable networks until very recently. Africa had far fewer cables and much less Internet coverage.
These precedents are crucial in understanding our current Internet. The Internet may seem wireless; but actually fiber-optic cables carry 95 to 99 percent of international data. Thinking about the history of infrastructure pushes us to look at the full spectrum of platform companies’ businesses. It turns out, for instance, that Google and Facebook are also infrastructure providers. Google partly owns 8.5 percent of all submarine cables.15 Just as the Eastern and Associated Telegraph Companies eventually expanded to Africa, so too are Facebook and Google: both companies intend to lay cables to Africa.16 Around a quarter to a third of Africans have internet access at present; by supplying the cables, Google and Facebook hope to increase the capacity of cables to Africa, lower the cost, and massively increase the market for their products.
Google is fully funding a cable from Portugal to South Africa via Nigeria. The company will name the cable Equiano, in honor of Olaudah Equiano, a Nigerian former slave who campaigned for the abolition of slavery in the eighteenth century. Equiano wrote about his experiences and travelled to London to push for the end of slavery.17 There is much irony in the name. The cable will land in South Africa, formerly a white dominion, and the site where Britain’s All-Red Route landed in 1902. An American-based company has appropriated the name of a former slave, while the cable itself represents an attempt by a Western company to appropriate provision of African communications.
More broadly, the cable ramps up competition between larger powers (the United States and China) over communications space. Chinese firm Huawei built around 70 percent of Africa’s 4G connections.18 Laying cables is part of a broader infrastructural competition over the supply of internet access to Africa. Beyond Africa, the Chinese government and Chinese companies are investing in 5G infrastructure while building international information networks through the news agency Xinhua, and a Belt and Road News Network (to accompany the Belt and Road Initiative’s other infrastructural projects). China aims to set the standards for 5G networks as a way to assert greater control over the next phase of global communications.
In the 1970s, Third World nations from Africa and Latin America called for a New International Information Order (later the New World Information and Communication Order).19 This was supposed to push back against Western dominance of news supply. It paid more attention to news firms such as news agencies than infrastructure. Now, however, African nations seem less concerned about China providing internet connectivity. Emeka Umejei from the American University of Nigeria noted in March 2019 that “most policymakers and politicians in Africa … don’t really care” about allegations that Huawei had installed listening devices in the African Union’s headquarters, a complex built by Chinese companies. Umejei called Africa “a pawn on the global chessboard in the ongoing geopolitical context.”20
China follows in a long tradition of states that see infrastructure and information as inextricably intertwined. These states invest in infrastructure for informational, geopolitical, and economic gains. The increasing contemporary attention to infrastructure parallels developments in the 1890s. Prior to that decade, most states were content with the submarine cable system and saw it as a neutral conduit of information. As international competition began to heat up between countries like Britain and Germany in the 1890s, both states started to see cables as the locus for growing geopolitical jockeying. Many states worried that cables were not neutral conduits of content. They feared, moreover, that states might subject cables to surveillance, that they might censor content, and that they might even cut cables in the event of a war.
A few decades later, these concerns led to infrastructure warfare. One of Britain’s first acts during World War I was to cut submarine cables connecting Germany to the world. In retaliation, German submarines devoted massive resources to cutting British cables throughout the war. From May 1915 to April 1917 (when the United States entered the war), the German Navy cut every cable starting from Britain, except those across the Atlantic. These were sophisticated efforts. On occasion, the Germans even used a rheostat to emit false electrical signals about where the break in a submarine cable had occurred, which made it harder to repair the cables swiftly.21 Cables were as much a part of the war as other weapons.
Internet infrastructure receives surprising little attention in the press and scholarly communities. Perhaps cables seem too far removed from our everyday experiences with wireless smartphones. But these cables make international communication possible and we ignore them at our peril. Information warfare is enabled by infrastructure, whether submarine cables a century ago or fiber-optic cables today.
Just as the history encourages us to look at infrastructures, it also encourages us to look at the structures enabling content dissemination. The history of the media industry should push us to pay attention to business structures as a crucial determinant of content.
3. Business structures are often more crucial than individual pieces of content.
The third historical pattern is that business structures are often more crucial than individual pieces of content. It is tempting to focus on the harm created by particular viral posts, but that virality is enabled by a few major companies who control the bottlenecks of information. Only 29 percent of Americans or Brits understand that their Facebook news feed is algorithmically organized; the most aware are the Finns at 39 percent.22 This control affords social media platforms huge power.
That power stems from the market dominance of platform and social media companies. Amazon, Apple, Alphabet (the parent company of Google and YouTube), Facebook (which also owns Instagram and WhatsApp), and Microsoft (owner of LinkedIn) together comprise one-seventh of the total value of the American stock market.23 That concentration of companies in a particular sector of the stock market is unprecedented.
However, business history can help us to understand how such circumstances affect content. For over a decade, business historians have been calling for scholars of management and entrepreneurship to take history seriously.24 This is no less true for the media business. It is notable that the runaway hit of 2019 on platforms was written by an emerita professor from Harvard Business School, Shoshana Zuboff. Zuboff argues that the companies accumulate data and are already using it to nudge our behavior. She calls this phenomenon “surveillance capitalism” because the companies surveil online behavior in order to monetize it. The ability to track people’s behavior across the Internet became key to the companies’ success.25 Some critics, like Evgeny Morozov, argue that Zuboff’s book mischaracterizes the capitalist aspect of the companies’ business model, which may be less effective in its targeting and advertising than it might seem.26
Business history offers several new ways to understand current problems. First, it pushes us to understand that bottlenecks have always existed in modern news delivery. Now it is Facebook, Google and co. But those companies’ role as a bottleneck for news resembles that of news agencies in the nineteenth and twentieth centuries. From the mid-nineteenth century, news agencies were similarly powerful. These were companies like Reuters, that used the new technology of submarine cables to gather news from around the world and telegraph it back home for newspapers to print. Because foreign correspondents and telegrams were so expensive, only a few news agencies existed. They became gatekeepers controlling the flow of information. News agencies possessed astonishing power. In 1926, 90 percent of all German newspapers had no correspondents abroad or in Berlin. They received all their national and international news through news agencies or syndicate services. It may now be algorithmic, but the problem of a few companies dominating news and determining how it is delivered is an old issue.
Ironically, news agencies have become more powerful in print media again over the last few decades. More and more newspapers have cut foreign correspondents, so more newspapers print wire stories than ever, even large newspapers like the Globe & Mail. On July 22, 2019, for example, the Globe & Mail front section included nine international stories; eight of them came from non-Canadian news agencies or the New York Times.27 This concern is long-standing. In 2008, journalist Nick Davies published a book criticizing British newspapers’ excessive reliance on news agencies for information.28
Second, a business history approach shows how ownership can affect overall directions in content. New business structures like vertical integration and cross-subsidies were able to create concentration and corresponding power in the news market. One key example of this in Weimar Germany was Alfred Hugenberg. Hugenberg began as a local bureaucrat, then moved into heavy industry in the Ruhr region of West Germany before starting to accumulate a media empire just before 1914. Unlike other newspaper magnates like William Randolph Hearst or Lord Northcliffe, Hugenberg succeeded by importing techniques of vertical integration from heavy industry firms like Krupp.
Hugenberg used vertical integration to incorporate all aspects of the newspaper business from paper to advertising. In 1916, he purchased the ailing publishing house, August Scherl, which published many leading newspapers, like Berliner Lokal-Anzeiger and Der Tag, and popular magazines, like Die Gartenlaube and Berliner Illustrierte Nachtausgabe. Hugenberg founded the advertising agency, Allgemeine Anzeigen GmbH (ALA), in 1917 and owned numerous paper companies. In 1927, Hugenberg purchased Universum-Film AG (UFA), which produced and distributed films and cinema news reels called Wochenschauen. UFA was a 1920s YouTube (without the user-generated content). At that time, cinema newsreels were a new and critical form of news consumption. Largely forgotten today, they ran before every film. Most newspaper readers and cinema goers probably had little idea that Hugenberg owned their entire media diet.
The hidden networks of Hugenberg’s media products extended to a news agency, Telegraph Union. This was a loss-making company that received cross-subsidies from other, more successful firms in Hugenberg’s portfolio. From the early 1920s, newspapers faced increasing financial issues (due to rising paper prices, hyperinflation, and increased fixed costs), and Hugenberg’s companies offered subsidies to small newspapers as long as they subscribed to Telegraph Union. Even ostensibly nonpartisan papers often unwittingly presented a nationalist take by printing news from Telegraph Union, particularly in the provinces. The agency’s increasing success polarized the supply of information.
Hugenberg shaped his media empire as a right-wing enterprise with no party affiliation, believing that readers would stop reading newspapers that too obviously pushed one political party or industrial sector. Instead, Hugenberg’s media enterprises supported antisocialist and nationalist politics in general. From 1920 onward, every editor working for Hugenberg’s Telegraph Union was contractually obliged to “campaign for the route of political and economic reconstruction of Germany without party-political or other ties on a national basis.”29
Telegraph Union exerted tremendous power by framing events and setting news agendas. That power would not generate political success: Hugenberg’s political party lost half its vote from 1928 to 1933 (from 14 to 7 percent). But these dynamics undermined the shared space for news within the increasingly febrile Weimar Republic and unintentionally laid the groundwork for the even more nationalist Nazis.
We tend to remember the Weimar Republic’s vibrant urban media culture, which was mostly liberal or left wing, but the business structures of Hugenberg’s media world were equally important. Similar problems plague our current analysis, where journalists and policy analysts still focus on celebrating the “Trump bump” in New York Times subscriptions and have only just started to understand the problems in local news beyond major urban centers. These analysts have not yet devoted sufficient energy to understanding the long-term trends, like those fostered by right-wing talk radio or other innovative conservative media initiatives and business structures. By contrast, historians like Nicole Hemmer and Brian Rosenwald are tracing the long-term dynamics of how conservative media activists and formats like talk radio might have been more important for explaining the rise of Trump than Fox News.30 And Jen Schradie’s work demonstrates that conservative activists have taken advantage of our new media environment more ably than groups on the left.31 These dynamics perhaps parallel Hugenberg’s successes (and maybe also his electoral failures because he was outmaneuvered by the further-right forces of the Nazi Party).
A focus on funding and business illuminates contemporary dynamics too. Many of the suggested reforms to social media companies are really about the companies’ business model. The companies optimize for engagement: they are content-agonistic. This means they prioritize content that generates engagement and more time spent on the site. Which in turn generates more advertising dollars. It does not matter if that content is extremist or cat videos. The companies are also incentivized not to investigate whether their content has problematic effects on users or, indeed, to reveal exactly how many people engage and with what intensity to which content. One obvious example is the President Donald Trump’s assertions of “conservative bias” from social media companies. The companies could publish investigations, which would almost certainly reveal that the claim is flawed. President Trump is highly unlikely to accept that finding. In August 2019, former Senator Jon Kyl, a Republican who represented Arizona, published a report commissioned by Facebook on the issue. Kyl’s short report drew from interviews with over 100 unnamed groups and individuals to enumerate conservative concerns;32 it focused on conservatives’ subjective experience of the platform without statistics published by Facebook itself. Facebook has not commissioned similar investigations for marginalized groups or even Democrats. The companies currently continue with models that optimize for engagement, no matter the externalities. Nicholas John has termed this “agnotology”: the counter-intuitive idea that the companies’ business model requires them to assert high engagement or effective algorithms but not to investigate the full effects or to reveal transparent numbers.33
The importance of ownership also extends to more conventional media products. Rupert Murdoch’s Fox News and newspaper outlets are an obvious example. Oligarchs and publishers loyal to Viktor Orbán have silenced dissenting voices by purchasing Hungarian media outlets. In November 2018, nearly 500 media companies were transferred to a non-profit foundation led by a publisher close to Orbán.34 We ignore newspapers, TV, and radio at our peril. Although their power is diminished, it remains vital.
If media history reminds us to look at business structures, the present shows how transnational those structures can be. Far-right news outlets like Rebel Media in Canada seem to be funded by the American anti-Muslim, far-right think tank, Middle East.35 And Rebel Media was at one point paying Tommy Robinson, a leading far-right figure in the United Kingdom who founded the English Defence League in 2009.36 There are also currently questions over a Saudi Arabian partial purchase of the Evening Standard, a London newspaper edited from 2017 to 2020 by the former Chancellor of the Exchequer, George Osborne.37 Meanwhile, Chinese media influence has a far reach, as one project (Chinfluence) is investigating in Eastern Europe. While most coverage appears to continue unaffected, Chinese ownership of Czech media led to much more positive coverage of China.38 The history of Hugenberg reminds us that we may not find the smoking gun of an owner telling journalists what to print; broader direction and ownership structures are enough. For tech companies too, business models explain much of the content we see. Alternative business models may solve more problems online than tinkering around the edges.
4. We need to design robust regulatory institutions and democracy-proof our solutions.
It is understandable that politicians worry in particular about elections and interference during campaigns, and many of the initiatives to counter disinformation focus on political consequences, such as the EU Code of Conduct for Disinformation, US proposals for an Honest Ads Act, or the Canadian Election Modernization Act. The German Network Enforcement Law (Netzwerkdurchsetzungsgesetz or NetzDG) enforces twenty-two statutes of German speech law online; it was passed swiftly before a German election in fall 2017 to show government action against social media companies.
However, the focus on the next election and the short-term can obscure the long-term consequences of regulatory action. Often the most important developments take years to understand. Talk radio in the United States is a good example; another is the unintended consequences of spoken radio regulation in Weimar Germany. Bureaucrats aimed to save democracy by increasing state supervision over content. This was meant to prevent seditious material that would bolster anti-democratic sentiment and actions. Ironically, however, these regulations ensured that the Nazis could far more swiftly co-opt radio content once they came to power in January 1933.39 Well-intentioned regulation had tragic, unintended consequences.
Weimar bureaucrats actively attempted to shape the media to save German democracy. They tried everything, ranging from subsidies to laws banning particular newspapers. A Law for the Protection of the Republic was passed in 1922; and while the Weimar Republic had press freedom, this legislation foresaw the restriction of freedom in exceptional circumstances. Nearly a decade later, in 1931, with rising violence on the streets, emergency decrees banned entire editions of newspapers for seditious content. There were 284 bans in total, including ninety-nine for Nazi papers and seventy-seven for Rote Fahne (the Communist newspaper) between 1930 and 1932.40
Officials also tried to withhold official government news from Alfred Hugenberg’s anti-republican newspapers, particularly Berliner Lokal-Anzeiger, Berliner Illustrierte Nachtausgabe, and Der Tag. In December 1929, the Prussian Ministry of the Interior decreed that it would stop supplying these three newspapers with official publications due to their “invidious and extremely provocative way” of attacking the government and form of the state.41 The Social Democratic Minister of the Interior, Carl Severing, hoped that removing official material from the “anti-state press … would lead without further ado to a corresponding reorientation of the reading public.”42 Other ministries (like the Finance Ministry) disagreed and found it “improbable” that readers would subscribe to different papers, just to get official news.43 In Weimar Germany at least, bans seemed to exert no measurable effect on readership. Hugenberg’s newspapers started to receive government news again in 1932, after bringing a court case on the matter.44
We see a similar debate now about banning various figures like Alex Jones from social media. Will it amplify their message or remove them from view? Will it stoke claims of “conservative bias”? Will bans change users’ habits or not? Multiple European countries like France and Germany have either enacted or are currently considering regulatory suggestions about enforcing bans on hate speech online. In the case of Germany’s law, NetzDG, a prominent AfD politician, Beatrix von Storch, had a social media post removed the day that the law came into force. This promoted considerable discussion amongst journalists and ironically amplified von Storch’s message, as well as giving prominence to the AfD’s assertions that they were being censored by both mainstream news outlets and social media.45 In fact, removing whole networks can be counterproductive by pushing them to migrate to another platform and amplifying their sense of victimization. Wholesale banning may be less effective on social media platforms than other strategies, such as banning small groups of users from online hate clusters (groups of users propagating hate speech).46
Other regulatory debates similarly focus on removal over other possible solutions. The European Union plans to introduce terrorist content regulation that will require social media companies to remove terrorist content within one hour. The regulation does not define terrorism and leaves it to member-states to do so.47 It is troubling if legislation allows leaders like Viktor Orbán to define terrorism as they please. A historical view reminds us that any media legislation has to stand in the long term. Some might like a hate speech law requiring removals under President Emmanuel Macron; but would they like it under a President Marine Le Pen?
Any productive approach to regulation should consider how to democracy-proof our systems. Institutional design is key here. Robust institutions would, for instance, consistently include civil society. They would bolster data security and privacy. They would also be designed not to lock in the current big players and shut down possibilities for further innovation.
In the United States, for example, campaign finance reform would likely prove more effective than other suggestions. This does not directly appear to address tech companies, but it would address their increasingly important role in campaigns. Both Democrats and Republicans now outsource communication to companies like Facebook. (Facebook embedded employees in the Trump campaign, for instance.)48 At the same time, campaign finance reform would address longer-standing issues of influence from billionaires and hidden campaigners, as discussed in other chapters in this volume. These reforms would affect all candidates and charge the Federal Elections Commission (FEC) with examining financial flows rather than content.
Other suggestions specifically for social media include regulating for transparency before intervening in content. A French proposal in May 2019, suggested the creation of a regulator who would enforce transparency and accountability from the largest social media companies. The idea is to create an ex ante regulator who will enable greater transparency from the companies and more involvement from civil society. The proposal followed a unique experiment where French civil servants were embedded at Facebook for several months.49 This regulator would also enable third-party access for researchers. Such proposals are less interventionist than many other suggestions and less appealing to many clamoring for the regulation of content. Such calls are particularly understandable from people who have suffered extensively from doxing or abuse online. But it is worth considering whether less interventionist solutions will better uphold democracy in the long run. It is also worth considering whether much of the abuse is enabled by the particular business models of social media and the lack of incentives to enforce their terms of service, which often already ban the behavior of abusive users.
One thing historians know is that humans are consistently terrible at predicting the future. We cannot foresee all the unintended consequences of our well-intentioned interventions. That does not mean we should do nothing, but it does warn us to democracy-proof our policy solutions. Or we might find ourselves undermining the very freedoms that we seek to protect.
5. Solutions must address the societal divisions exploited on social media.
The seeds of authoritarianism need fertile soil to grow; if we don’t address underlying economic and social issues, communications cannot obscure discontent forever. It would be an extreme oversimplification, for example, to attribute the rise of the Nazis to media strategies. The Great Depression, political unrest, discontent stoked after the loss of World War I and the Versailles Treaty, and elite machinations all played essential roles.50
Media amplified certain aspects of discontent and contributed to systemic instability. The continual coverage of scandals by papers across the political spectrum conveyed a sense of a democratic system that was not working. Historian Corey Ross has argued that German interwar obsessions with propaganda undermined the Weimar Republic “not only by nourishing right-wing notions of an authoritarian Volksgemeinschaft, but also by eroding democratic conceptualizations of public opinion across the political spectrum.”51 These attitudes mattered, but political behavior also dovetailed with people’s lived experiences of hyperinflation, unemployment, and street violence.
Media effects research over the past century warns us to beware of simple assumptions that equate exposure to media with political outcomes. So does historical research on the Weimar Republic. Bernhard Fulda examined a small town in Germany with one – right-wing – newspaper, which recommended its readers vote one way in a referendum in the mid-1920s.52 The majority of the town voted the other way. Another study has found that Hitler’s speeches appeared to have negligible effect on how people voted (other than possibly in the presidential election of 1932). This suggests that media coverage of Hitler’s charismatic speeches was less influential than scholars had previously assumed.53 Many other economic and social factors clearly shaped voter behavior. This does not mean that media do not matter. It means that we must be careful to over-ascribe efficacy to individual pieces of content. The same is true for social media.54
Just as media in the Weimar Republic exploited or deepened extant social divisions, social media today often does the same. What has changed is the algorithmic and microtargeted delivery of news. Algorithms amplify particular pieces of content to increase engagement; Russian trolls, for example, have used this to their advantage by focusing on stoking controversy around issues such as Black Lives Matter or vaccination. People are most likely to share material online that angers them. The negative emotion of anger decreases our analytical functions, so we are more likely to believe the material; we are also more likely to repost it. As social media companies optimize for content that increases engagement, their algorithms may supply more material that angers us, inspiring sharing and engagement.55 The algorithmic bias toward anger is new; our anger-inspired analytical biases are not. Social media may amplify anger, but that anger also stems from real-world experiences of current conditions. As we continue to debate how best to address legacy and social media, we should not focus on those problems to the exclusion of others. Sometimes, media scholars are the people best placed to argue that other policy areas matter more. If we do not address pressing issues like growing inequality and climate change, improved social media communication will not stem discontent.
Conclusion
Over the past decade, I worked on a book about how Germany tried – and almost succeeded – in its attempts to control world communications from 1900 to 1945.56 Amongst other things, I explain how Germany’s democracy, with its vibrant media landscape, could descend into an authoritarian, Nazi regime spreading anti-Semitic, homophobic, and racist content around the world.
While I was writing this book, the present caught up with history. Far-right groups in Germany and around the world revived Nazi terminology like Lügenpresse (lying press) or Systempresse (system press) to decry the media. News was falsified for political and economic purposes. Minority groups were targeted and blamed for societal ills that they did not cause. As with radio, internet technologies designed with utopian aims have become tools for demagogues and dictators.
As these events unfolded, scholars tried to combat erroneous assertions of novelty. As Michael Schudson and Barbie Zelizer wrote in 2018, “To act as if today’s fake news environment is fundamentally different from that of earlier times misreads how entrenched fake news and broader attitudes toward fakery have been.”57 Attitudes toward fakery have changed over time and depending upon the medium. Andie Tucher has shown that in the late nineteenth and early twentieth centuries, faking in photography was prized as a way to make something appear more real.58 John Maxwell Hamilton and I have explored different forms of falsification in the history of news: faking for political purposes, both domestic and international; and faking for economic purposes, either to increase a newspaper’s circulation or to boost a product.59
What I have discussed in this chapter is not the content itself, but rather the structural conditions enabling falsification or disinformation. First, disinformation is also an international relations problem. Second, physical infrastructure matters. Third, business structures are more important than individual pieces of content. Fourth, robust regulatory institutions must take a long-term view that balances between protecting freedom of expression and protecting democracy. Fifth, media exploit extant societal divisions.
Five years ago, the question was if we would regulate social media. Now the questions are when and how. That development is a good one. But for regulation to protect democracy, we should also consider the questions raised by broader historical patterns.
Open access