Advances in digital technology are dramatically transforming our lives, often for the better. Social networking, cloud computing, data analysis, cryptography, and artificial intelligence are not only making us more productive, efficient, and connected; they also promise solutions to thorny social problems, from poverty and state oppression to climate change and even human mortality.
As a torrent of scandals has revealed, digital progress also has a dark side. Technology can exacerbate social divisions, facilitate the execution of crime and violence, and create staggering inequalities. Philosophers have heeded the call, joining computer scientists, journalists, and lawyers to diagnose these challenges and illustrate the ethical stakes of technological change. Emerging research offers powerful accounts of how core values like privacy and fairness apply to product design, how risk and responsibility for technological harm should be distributed, and how technology firms can be better regulated.
But the incipient ethics of technology subfield has so far had little to say about the challenges that technology poses to another key value: the value of democracy. On the view that I favor, democracy requires that major social conditions result from collectively authorized procedures that allow each affected person fair opportunities for informed influence. Digital technology production threatens this ideal insofar as it privatizes or automates decisions about major social conditions, creates or widens political inequalities, and dominates fora for public deliberation. Though by no means the only or even the most urgent challenges for the ethics of technology, these challenges are serious. More importantly, leading ways of appraising technological developments tend to obscure these challenges from view.
In what follows, I briefly survey how four paradigms in technology ethics—substantive justice, consumer welfarism, democratic minimalism, and workplace democracy—overlook a critical set of concerns with a democratic cast. I then show how taking democracy seriously might generate distinctive principles of conduct for technology producers.
Four Themes in Technology Ethics
Substantive justice | Recent philosophical work in this area has attended especially to questions of existential risk, privacy, discrimination, and responsibility. These contributions have advanced our understanding of the hazards that artificial intelligence poses to human safety and existence. They have explored the right to privacy in online interaction and the threats to that right from data collection and management. They have revealed how algorithms that draw on data from socially stratified societies may reproduce or intensify forms of social discrimination. And they have shown how the safety risks of autonomous vehicles can be fairly distributed, as well as who bears moral responsibility when autonomous systems cause harm.
A common thread knits these perspectives together. Each takes concern with whether and how digital technology can be rendered safe and fair—questions about the substantive rightness or goodness of technological developments. These are obviously important questions, and compelling answers can be derived with the tools of moral philosophy. But if we consider the ethics of technology from the standpoint of political philosophy, these questions appear to miss a key dimension of ethical appraisal.
In evaluating social conditions, political philosophy often distinguishes principles of justice (standards for evaluating social conditions) from principles of legitimacy (standards for evaluating the processes that create these conditions). While principles of justice give us tools to answer “what the rules should be,” principles of legitimacy give us tools to evaluate “who makes the rules.”
In my assessment, the existing philosophical literature on the ethics of technology has focused on questions of the former kind at the expense of the latter. It is one thing to appraise the substantive properties of technological products, and another thing to appraise how control over product development, design, and use is distributed. To ignore questions of legitimacy is to ignore questions of power. And such questions are especially important in this domain for two reasons.
First, digital technology production lies largely with private firms with hierarchical structures of internal governance. As hierarchical corporations, technology firms afford limited accountability to those who use their services. Amazon customers can rate products, badger customer service representatives, and boycott, but they can’t vote in company decisions about its market power, workplace policies, or environmental impact. And as private organizations, technology firms are not directly accountable to third parties. This lack of accountability matters because of the externalities that digital technology products generate. I might refuse internet access or never use Google’s search engine. But because Google effectively enjoys a monopoly on internet search, it controls the flow of information and enjoys substantial influence over which ideas prevail in my society.
Second, technological developments threaten to reshape the exercise of political authority. In some areas, big data and AI offer possibilities for radically expanding the powers of state officials. Predictive policing and facial recognition are already revolutionizing the field of law enforcement. In other areas, digital innovation promises to pry control away from states—or out of human hands altogether. Blockchain technology can be used to replace government functions in currency provision and contract assurance. Artificial intelligence is starting to assist or even commandeer legislative, administrative, and judicial decisions, potentially drawing authority away from human citizens, officials, and judges.
Consumer welfare | Some scholars have recently taken an interest in the outsized power of technology platforms, systems of digital infrastructure—like Netflix, eBay, and Twitter—that provide a base or marketplace for other applications or interactions. Platforms thrive on network effects, in the sense that the value of the platform to its users increases the more individuals join it. Because the cost of exiting the platform also increases the more individuals join it, successful platforms are often resistant to competitive pressure.
The anti-competitive tendencies of digital platforms have led several legal theorists to advocate for the application of “antitrust” laws to these firms. These theorists appeal to American jurisprudence of the progressive era, which saw industrial consolidation as a threat to consumer welfare. In my view, the arguments to this effect are deeply compelling. Platforms in some areas of the digital economy do indeed impose objectionable costs on consumer welfare, by limiting choice and driving up prices. But I also think that arguments grounded in a concern for consumer welfare fail to appreciate other and potentially more serious objections to monopoly.
Critics of Amazon rarely focus on choice or price. Rather, they decry the dominating power that Amazon wields over its sellers and employees, its ability to bully state and municipal governments, its contributions to deforestation and hyperconsumerism, and its predatory displacement of smaller businesses and traditional retail establishments. Consumer welfare may be a compelling legal objection, but it’s a weak moral objection in light of the interests that consumers have beyond consumption and the interests of non-consumers who must suffer the externalities of monopoly practices. In short, a preoccupation with antitrust law perilously overlooks more central objections to monopoly, objections that can’t be answered without engaging democratic theory.
Democratic minimalism | In fairness, democratic critiques of digital technology are increasingly common. In particular, critics have worried about foreign infiltration of electronic voting systems, the deployment of online “bots” to lobby public officials, advanced propaganda techniques (e.g., “fake news” campaigns), and the ways in which social networks encourage partisan division.
Each of these concerns is well taken. The primary limitation of such critiques is their tendency to portray electoral integrity as the one and only principle of democratic legitimacy. They imply that as long as technology steers clear of interference with the selection of candidates for public office, technology respects democracy. Such a position might draw support from minimalist theories of democracy. These theories identify democracy’s value in its ability to reduce gross injustice by periodically forcing governors to stand for election. Whatever the virtues of minimalist theories of democracy, however, it should be clear that these perspectives are unable to account for the concerns I’ve raised above regarding privatization, automation, hierarchy, and the like.
Workplace democracy | More serious engagements with democratic theory have recently explored the internal governance of technology firms. Noting that such firms have come to enjoy overwhelming control over the lives of their employees, they suggest that employees ought to enjoy greater voice in internal governance decisions. Though I see many good reasons to democratize the workplace, I believe workplace democracy is neither necessary nor sufficient to legitimate the power of technology firms.
If the basis of our objection to the power of technology firms is that they dominate their employees, a compelling reply is that this domination is primarily the result of background conditions that unjustly limit employees’ bargaining power. The proponent of workplace democracy either needs to show that corporate hierarchies constitute a distinctive injustice apart from these background conditions, or else that reforming corporate governance is simply a more feasible possibility than limiting monopolies and ensuring that workers enjoy a decent social minimum that enhances their bargaining position.
If the basis of our objection to the power of technology firms is that they dominate society, it’s hard to see how reforms to their internal governance could be a sufficient corrective. Making Amazon more internally democratic might give its employees greater control over their lives, but it wouldn’t do much for customers, competitors, and local communities whose lives are also deeply affected by Amazon’s decisions.
What Makes Democracy Valuable?
To justify and explain digital technology’s democratic deficits, we need a more robust theory of what makes democracy valuable and how that value applies to different agents.
Let “democracy” refer to a process of collective decision-making that treats members of the collective as equals in some way. In this bare formulation, the concept admits a wide variety of applications (such as nation-states, associations, and firms) and configurations, such as direct and representative forms of democracy, as well as mixed constitutions that split power between elected and unelected bodies. Important for my purposes is that the democratic ideal also provides the dominant conception of political legitimacy—that is, a basis for evaluating the normative standing of coercive decisions. (I ultimately suggest that standards of political legitimacy can devolve to non-coercive decisions in certain cases.) Such decisions that issue from democratic procedures enjoy a certain moral authority that non-democratic decisions lack. Democracy gives subjects of coercive decisions strong reasons to comply even when they disagree with the content of these decisions.
Precisely what lends democratic decision-making this authority provokes significant disagreement. But this debate typically homes in on four values: collective self-determination, political equality, deliberative justification, and substantive reliability. Other forms of rule might exhibit one or more of these desirable features to some extent, but only the democratic ideal can capture all of them at once.
Collective self-determination refers to the idea that the “makers” and “matter” of the laws, the rulers and the ruled, are one and the same. Democracy is valuable in part because it affords subjects of authority a measure of control over their common affairs. Forms of authoritarianism—dictatorships, theocracies, colonial domination, and so on—serve as natural contrasts. Which aspects of society must be democratically controlled in order for citizens to enjoy collective self-determination remains undertheorized. A common position holds that collective self-determination grounds a claim to external sovereignty—a (possibly defeasible) right to exclude foreign agents from interfering in domestic affairs. A more demanding position holds that collective self-determination also requires internal sovereignty. Citizens can’t be said to enjoy collective self-determination unless they can effectively control the basic structure of their society, i.e., the major social institutions that create and distribute basic rights, duties, and opportunities.
Political equality refers to conditions that allow citizens to enjoy this collective control on equal terms. A foundational premise of political morality holds that adult human persons are entitled to equal respect and concern. Several recent prominent papers, including by Niko Kolodny and Daniel Viehoff, argue that democracy is the only decision-making procedure that fully respects subjects as equal moral persons. Attempts to mark off some people as more qualified to rule than others must draw distinctions between the more competent and the less competent. But even if we could identify criteria for competence in the abstract, attempts to apply such criteria in practice require making discriminatory judgments or (what David Estlund calls) “invidious comparisons” in order to sort people. Thus, any decision-making procedure that restricts political authority to certain people fails to respect the equal moral status of all. Democracy, on the other hand, consecrates moral equality by refusing to draw these kinds of distinctions. Democratic procedures uniquely affirm the bedrock principle of equal respect for persons. A critical implication of this position is that citizens ought to enjoy equal opportunities to influence their common affairs. Equally-weighted votes are an important mechanism for assuring this condition, but so are fair opportunities for influence over public discourse and the political agenda.
Deliberative justification prevails when decisions follow an open process of reflective reason-giving rather than a simple tallying of people’s preexisting preferences. Deliberation helps to protect against the possibility that outcomes will reflect ignorance or prejudice. That decisions can be defended with reference to sound arguments also enhances their justifiability to those who fall subject to them.
Finally, substantive reliability characterizes a system of rule that consistently generates high quality outcomes—ones that promote the common good and satisfy principles of justice. Democratic governance is thought to satisfy this criterion partly due to the epistemic properties of deliberation, which exposes ideas to scrutiny and screens out those political positions that can’t be broadly accepted. Democracy also tends to generate high-quality outcomes because of the ways that it aggregates the wisdom of the multitude, appoints competent officials through institutions of representation, and reproduces itself stably over time.
Democracy’s critics counter that democratic processes are no guarantee against unjust outcomes; in many cases, alternative forms of rule might yield better results. Indeed, democratic majorities may choose to elect tyrants, discount future generations, or violate the rights of minorities. This criticism loses much of its force if we understand the democratic ideal as one that contains built-in constraints and scope conditions. Constitutional limitations on majority rule are consistent with the value of democracy because they preserve the very foundations of democratic governance. On this reading, unbridled majoritarianism is a perversion, rather than an implication, of the democratic ideal. Similarly, according to this perspective, the value of democracy doesn’t require equally distributed collective control over all aspects of social life. The way it applies is far more nuanced.
Democratic legitimacy applies principally to the regulation of the basic structure of society—the set of institutions that establish the ground rules for social interaction and determine fundamental rights, duties, and opportunities. This entails that democracy provides the appropriate standard of legitimacy for the design, administration, and enforcement of laws. Democratic legitimacy doesn’t ordinarily apply to the internal governance of private associations and personal relationships. To count as legitimate, firms, religious societies, and clubs needn’t necessarily run themselves by democratic procedures. If a society’s basic structure is already regulated by democratic procedures, why shouldn’t citizens be free to form and join associations with diverse forms of internal organization? But democratic legitimacy does impose a set of demands on the conduct of private associations. As I’ll elaborate, it requires that private associations respect the scope and foundations of democratic authority—requirements that impinge on technology producers in key ways.
Democratic Duties
When the basic structure of society is in fact democratically legitimate—i.e., responsive to mechanisms of collective control that afford each member of the collective a fair opportunity for informed influence—private associations are simply required not to supplant or circumvent them.
For example, consider cryptocurrencies such as Bitcoin or Libra, which use innovations in cryptography and distributed ledger technology (e.g., blockchain) to provide a means of exchange that is neither authorized nor controlled by states. Arguably, however, monetary policy is a component of the basic structure, as it has a pervasive effect on the nature and distribution of fundamental rights, duties, and opportunities. Moreover, unregulated currencies encourage economic inequality and criminal activity. When a state’s monetary policy satisfies the criteria of democratic legitimacy, my view suggests that private firms are duty-bound to respect the scope and foundations of this authority. Bringing cryptocurrencies to market under these conditions supplants and circumvents the legitimate authority of citizens to regulate core matters of collective concern together, on equal terms.
Note that to discourage private agents from supplanting or circumventing legitimate institutions isn’t to deny them opportunities to contest the outcomes of these institutions or even to engage in civil disobedience. Nor does this injunction prohibit private agents from supplementing democratically authorized policies, insofar as this can be done in ways that don’t threaten legitimate public institutions.
Moreover, when the basic structure is in fact democratically legitimate, private agents are not required to organize themselves democratically, nor are they required to give third parties a voice in internal decisions.
Things change, however, when democratic institutions are weak or absent. When the state fails to regulate the basic structure (or components thereof), or fails to ensure political equality, I propose that principles of democratic legitimacy can devolve onto private agents in certain ways. Namely, private agents incur (1) duties not to exploit the democratic deficit, (2) duties to facilitate transitions to legitimate democratic governance, and (3) duties to embody democratic constraints in their internal conduct.
Exploitation of the democratic deficit occurs when firms take undue advantage of political inequality, by leveraging their economic clout to quash regulation or to extract concessions from governments. Exploitation also occurs when firms operate like public utilities by providing essential infrastructure for commerce or communication—and yet constrain usage or access on biased or arbitrary grounds.
By facilitating transitions, I mean that when private agents come to enjoy power over elements of the basic structure, they ought to treat this power as temporary stewardship. For instance, if autonomous vehicle manufacturers find themselves in the position of setting policies about who lives and dies at the hands of their products, they shouldn’t seek to maintain this power permanently, but rather to cede it to democratically-authorized public bodies or encourage the creation of such bodies if they do not exist.
Private agents can embody democratic constraints in two main ways. One is to democratize internal governance by reducing inequalities in power among different categories of stakeholders. That is, when undemocratic government limits the political power of individual citizens, the stronger the claim may be for a voice in the governance of voluntary associations. Similarly, when undemocratic government leads to monopolies that dominate the lives of employees, customers, and nonconsenting third parties, the stronger the argument for a firm’s internal democratization. Though facially similar, these recommendations go beyond conventional workplace democracy arguments by explicitly treating internal democratization as a second-best option and by extending opportunities for influence to other stakeholders besides workers.
A second way of embodying democratic constraints may be required when platforms serve as gatekeepers to essential public facilities. For instance, the more Facebook and Twitter come to serve as gatekeepers for the public sphere, the more important it is that their policies regarding access and content moderation can be publicly justified and contested.
Discharging any of the duties I’ve proposed requires contextual judgments about the nature of the technology and the extent to which a democratic deficit exists. Does the technology in question engage questions of fundamental rights, duties, or opportunities? Is the technology in question supplementing democratic authority or supplanting it? Does the society in question have unjustifiable levels of political inequality? Is the state unable or unwilling to regulate the digital economy? More philosophical work is required to help establish criteria for making these judgments, such as what opportunities are “fundamental” and what forms of political inequality are “unjustifiable.” Making these judgments also depends crucially on the empirical investigation of actual conditions, in order to assess the capabilities of particular technologies or the functional limitations of particular governments. In other words, the democratic ethics of technology must be interdisciplinary through and through.
The foregoing discussion leaves much for further elaboration and debate. But two things, I hope, are now clear: The value of democracy makes a variety of demands on technology producers that go far beyond electoral noninterference, and philosophy can do more to help us understand the nature of these demands.
Photo by Today Testing
Ted Lechterman
Ted Lechterman(@tlechtable)is a postdoc at the Hertie School of Governance in Berlin. His current research addresses how the value of democracy applies to emergent economic practices. An earlier version of this piece was presented at the Centre for Advanced Studies Justitia Amplificata annual conference at Goethe University Frankfurt in 2019.
