2025-12-01 - The NFC Research Archive - A Critique of the Attention Economy: Flourishing, Autonomy, and Corruption

Benjamin H. Schwartz

A dissertation submitted to the faculty at the University of North Carolina at Chapel Hill in

partial fulfillment of the requirements for the degree of Doctor of Philosophy in the Department

of Philosophy in the College of Arts and Sciences.

Chapel Hill

2025

Approved by:

Luc Bovens

Daniel Muñoz

Douglas MacKay

Mariska Leunissen

Clinton Castro

Benjamin H. Schwartz

iii

ABSTRACT

Benjamin H. Schwartz: A Critique of the Attention Economy: Flourishing, Autonomy, and

Corruption

(Under the direction of Luc Bovens)

The attention economy is a widespread market characterized by two types of transactions:

people exchange their attention for access to new media services; these companies then sell our

attention, and the information they have gathered by means of it, to advertisers. My dissertation

argues that the attention economy is an impediment to human flourishing—that it is deeply at

odds with intimate relationships, attentional virtues, autonomy, and valuing attention properly.

First, I argue that dating apps, an attention economy service, undermine ways of thinking

central to forming and maintaining committed romantic relationships. Their choice structure

fosters a people-shopping mindset of exit and optimization. This economic way of thinking is

morally degrading in the context of intimate relationships. Moreover, dating apps have

diminished the availability of other modalities of dating and have created a collective action

problem.

Second, I explicate how sustained attention is inhibited by the attention economy and I

argue that this capacity is essential to virtue and human flourishing. In particular, the attention

economy inhibits the cultivation of the specific virtues of aloneness and stillness. These virtues

have inherent value and are essential for achieving other valuable goods, such as self-knowledge

and social commitments. Social media also engenders positional competition for inherently

scarce, attention-based goods like popularity.

Third, I argue that the attention economy infringes upon one’s autonomy by significantly

interfering with control over sustaining and directing one’s attention. The attention economy is

pervasive, practically necessary, and hidden. It objectionably interferes with people’s secondorder preference satisfaction and their internal preference formation. And, one’s attention-based

social obligations are at odds with the obligation to protect one’s autonomy from the attention

economy.

Fourth, I argue that the attention economy expresses and promotes corrupted attitudes

about attention, namely, that attention is instrumentally valuable, commensurable with market

goods, and valuable solely for its quantity. I develop a threshold view of when this corruption

occurs for attention and defend my view from Brennan & Jaworski’s critique of the argument

that markets express corrupted attitudes.

To my parents

ACKNOWLEDGEMENTS

This was an exceptionally demanding project that could not have been completed without

the assistance and generosity of my wonderful support network. I am very grateful to all of you.

I’d like to begin by thanking Luc Bovens, my dissertation chair, for being there with me every

step of the way. I cannot thank you enough for your insightful comments, engaging

conversations, kind and helpful advice, and for ensuring that I always felt supported. Thank you

to my committee members, Daniel Muñoz and Doug Mackay, for these crucial types of support,

including professional guidance, assistance with framing my ideas, and many fruitful

conversations. Thanks also to my two other readers—Mariska Leunissen and Clinton Castro—

for their helpful and insightful discussion of my work.

I am grateful to UNC Philosophy’s faculty and staff for providing such a welcoming and

collegial place for me to hone my philosophical skills. Thank you to my fellow UNC graduate

students, including Z Quanbeck, for being insightful and supportive philosophical interlocutors.

Thank you to James Lesher (“Dr. L”) for inspiring my love of Ancient Greek philosophy, as well

as for encouraging me to pursue graduate studies. I’d also like to thank the people I worked with

at the University of Washington, including professors John Manchak, David Keyt, and Andrea

Woody. Barbara Mack was an amazing graduate coordinator and I am grateful for her kindness.

I have been blessed with a large and loving family, and I couldn’t have finished this

project without everything you’ve done for me. Thank you, everyone! I am grateful to my

mom’s siblings, who have supported me and her with love, kindness, and constant aid over the

past few difficult years. I would also like to thank my friends, whose encouragement was a

vii

source of strength. Thank you to Kari and Schnitzel for your love, emotional support, and a

home base. Finally, it is impossible to express how essential my parents were to this project and

to who I am. They have been a lifelong source of intellectual and emotional support and

guidance; they are models for how to treat people with love, kindness, and respect. Thank you

and I love you.

viii

TABLE OF CONTENTS

LIST OF ABBREVIATIONS..........................................................................................................x

INTRODUCTION ...........................................................................................................................1

I. Prologue ...................................................................................................................1

II. Overview..................................................................................................................4

III. Remarks ...................................................................................................................7

References..........................................................................................................................11

CHAPTER ONE: DATING APPS AND THE LIMITS OF MARKETS.....................................12

I. Introduction............................................................................................................12

II. A Mindset at Odds with Committed Relationships ...............................................14

III. Problematic Norm Changes: People-Shopping .....................................................24

IV. Dating Apps and Technology-Induced Change.....................................................32

V. Conclusion .............................................................................................................40

References..........................................................................................................................41

CHAPTER TWO: ATTENTION, VIRTUE, AND THE ATTENTION ECONOMY .................43

I. Introduction............................................................................................................43

II. Sustained Attention and Virtue..............................................................................45

III. AE and Positional Goods.......................................................................................60

IV. Conclusion .............................................................................................................70

References..........................................................................................................................71

CHAPTER THREE: AUTONOMY AND THE ATTENTION ECONOMY...............................73

I. Introduction............................................................................................................73

II. Why AE violates our Autonomy............................................................................74

III. AE and Competing Obligations.............................................................................92

IV. Conclusion .............................................................................................................98

References..........................................................................................................................99

CHAPTER FOUR: CORRUPTION AND THE ATTENTION ECONOMY.............................101

I. Introduction..........................................................................................................101

II. Attention and the Corruption Objection ..............................................................102

III. A Response to Brennan and Jaworski’s Objection..............................................115

IV. Conclusion ...........................................................................................................125

References........................................................................................................................127

LIST OF ABBREVIATIONS

AE the attention economy

B&J Brennan and Jaworski

C&P Castro and Pham

EWOT the economic way of thinking

INTRODUCTION

I. Prologue

Smartphones, social media, and their associated apps are ubiquitous. In 2024, five major

U.S. technology companies had a collective stock value of $9 trillion, slightly less than 20 % of

the entire S&P 500.

1 Moreover, between 2011 and 2024, the percentage of U.S. adults who own

a smartphone nearly tripled from 35 to 91% (Pew Research, 2024). At first glance, these

products and services provide many apparent benefits to their users, including entertainment,

information, social connections, and a faster way to communicate. These technologies promise to

(say) help isolated individuals become more socially engaged by keeping up with old friends.

Also, people seem to freely choose to engage with these services. This squares with a

general pro-market argument. First, markets promote a preference satisfaction version of wellbeing. People trade their attention for these services because that’s what they prefer; both parties

to this voluntary exchange are mutually “better off” in terms of preference satisfaction. Second,

markets respect freedom. In this case, we should be allowed to trade our attention for internet

services; to disallow it would violate a version of freedom—the freedom to do what one wishes

with the things one owns (cf. Nozick, 1974). Is that enough to alleviate any moral concerns?

Apparently not. Recently, there has been a widespread sense that something is morally

problematic about smart phones, social media, and some of the companies that profit off these

1 These companies are “Meta…; Amazon; Apple; Netflix; and Alphabet”.

https://www.investopedia.com/terms/f/faang-stocks.asp

technologies. For one thing, we are in the midst of mental health and loneliness crises, especially

for the younger generations, that coincided with the rise of smartphones. (See Castro and Pham

(C&P), 2020: 3-4.) Relatedly, there have been multiple recent state-level laws and lawsuits

targeting key technology companies such as Facebook (Meta).2 This is on top of myriad news

articles, op-eds, popular books, etc. that express deep concern with the effect of new technology

on our lives (e.g., Cal Newport’s Digital Minimalism). Moreover, many television series and

films, such as Black Mirror and the movie Her, explore dystopian themes regarding our current

and near-future relationship to technology. Finally, sparked by pervasive concerns about the

corrosive impact of “low-quality” online and social media content “on a person’s mental or

intellectual state”, “the Oxford Word of the Year for 2024 [was] ‘brain rot’”.

This raises the question of whether there is a serious moral problem with these

technologies and (if so) what exactly it is. Importantly, I believe that the attention economy

(AE)—the economic system underlying recent digital technologies—is a significant part of the

moral problem, and in what follows, I hope to show why. A key aim of mine will be to help

clarify and deepen our understanding of what is morally objectionable about AE. But first, we

need to understand what AE is.

For this purpose, I will roughly follow Castro and Pham’s (2020) characterization of the

attention economy. They define AE as: “the economic market in new media

services…constituted by two types of transactions: [1] consumers give new media developers

their literal attention in exchange for a service… [2] developers auction off consumer attention to

advertisers” (C&P, 2020: 2, my numbering). Basically, people exchange their attention for

2 See, e.g., https://www.nytimes.com/2023/10/12/technology/tech-children-kids-laws.html?smid=pc-thedaily

3 https://corp.oup.com/news/brain-rot-named-oxford-word-of-the-year-2024/

access to a new media service or product. Companies then sell our attention, and the data it

generates, to advertisers. Expanding upon this characterization, in the first type of transaction,

our attention is used as currency (a means of exchange). In the second type, our attention is both

the company’s product and an input resource (due to the information it produces)—AE

companies generate revenue by capturing and selling our attention.

Some paradigm cases of new media companies that engage in these markets are internet

applications such as Facebook, Instagram, Twitter (X), TikTok, etc. These new media services

are located and delivered via the internet and are often accessed via smartphones. Moreover, AE

differs from previous advertising markets because new media services can “absorb and respond

to information about consumers in real time” (C&P, 2020: 1). This tends to make the services

and content they offer, along with their ads, much more specifically tailored and attentiongrabbing to a user of their site. For instance, YouTube’s algorithm, based on its analysis of the

user’s past behavior, might show two users from the same town (or even household) vastly

different content and ads in real-time; traditional television, for example, does not make such

disparate and fine-grained distinctions between these types of individual viewers (C&P, 2020: 1).

There are other important differences between AE and previous advertising markets. As I

argue in Chapter Two, AE substantially interferes with people’s capacity for sustained attention.

Moreover, as I argue in Chapter Three, AE is far more pervasive, practically necessary, and

hidden from people’s awareness than previous iterations of markets in attention.

While AE products and services are sometimes “free” in the sense that consumers do not

exchange money for accessing and using them, the scope of AE goes beyond these services and

products. As I explicate in Chapter Four, even some companies like Hulu that charge a monetary

fee still generate revenue through product placement as well as harvesting and selling user data.

Both of these practices incentivize capturing people’s attention, as well as reduce the cost of their

users’ subscription fees. Hence, for these companies, attention is still a product and input

resource, and also a means of exchange (since it lowers the monetary cost of user access).

II. Overview

My dissertation is structured around a handful of pressing questions in the ethics of

technology. Centrally, is AE helping us to live well? Is AE serving us or are we serving it?

Relatedly, is attention a commodity that should be bought and sold in a market?

Some critics of AE have argued that social media and smartphones cause addiction (e.g.,

Bhargava and Velasquez, 2021) and poor mental health (e.g., Haidt, 2024). However, I believe

that these are only aspects of a wider moral problem. Instead, I broaden the critique by arguing

that the attention economy is an impediment to human flourishing. I incorporate virtue-based

objections to markets to argue that AE is deeply at odds with intimate relationships, the crucial

virtues that involve attention, autonomy, and appropriate attitudes toward attention and its value.

Each chapter of my dissertation examines the intersection of AE with important aspects

of flourishing. I begin by critically analyzing dating apps. Since these apps are a key instance of

AE, and illustrate many of the moral problems with it, this chapter serves as a wedge for a more

general critique of AE in the subsequent chapters.

In Chapter One, I explicate how dating apps, particularly their market features, affect

ways of thinking that are crucial to maintaining long-term commitments and valuing intimate

relationships. In Chapter Two, I analyze the connection between AE and people’s attentionbased capacities and virtues, as well as how positional competition on social media impacts our

attitudes and relationships. In addition to the inherent value of autonomy, I assume that a well-

developed capacity for autonomous agency helps people to flourish. With this in mind, in

Chapter Three, I critically assess the connection between AE and autonomy. Finally, in Chapter

Four, I examine the symbolic meaning of this widespread market in attention (i.e., AE) and the

degrading impact this may have on our attitudes and values. I will now provide an overview of

my main positions and arguments in each of these chapters.

In Chapter One (“Dating Apps and the Limits of Markets”), I argue that dating apps, an

AE service, undermine ways of thinking that play a crucial role in long-term commitments and

valuing others appropriately in intimate relationships. First, I argue that the new choice structure

of dating apps, underpinned by their AE model, fosters a market mindset of exit and optimization

that is at odds with forming and maintaining committed relationships. Second, I expand upon

why dating apps foster a people-shopping mentality, and I argue that this economic way of

thinking—e.g., transactional attitudes and the commensuration of goods—is morally degrading

in the context of intimate relationships.

If so, why do people continue to use dating apps (and other AE services)? I argue that, by

initially appearing to add more choices for how to date, the widespread adoption of dating apps

diminished the availability of other modalities (e.g., in-person connections) and led to a difficult

collective action problem. I conclude by examining how to alleviate these downsides.

In Chapter Two (“Attention, Virtue, and the Attention Economy”), I develop an account

of how key forms of attention (especially sustained attention), which are inhibited by the

attention economy, are essential to virtue and human flourishing. Moreover, AE, paired with

smart phones, inhibits our ability to develop and maintain the specific virtues of aloneness and

stillness (along with a related tolerance for loneliness and boredom). In addition to having

inherent value, I argue that the virtues of aloneness and stillness are needed for achieving other

valuable goods, such as self-knowledge, completing long-term projects, and maintaining

committed relationships. Finally, I present and explicate an additional reason why AE aversely

impacts people’s ability to flourish—namely, social media relies upon and engenders an ongoing

positional competition for inherently scarce, attention-based goods like popularity.

In Chapter Three (“Autonomy and the Attention Economy”), I argue that AE

substantially violates our autonomy. First, I argue that because control over sustaining and

directing our attention is central to being an autonomous agent, AE’s significant interference

with this capacity is a deeper autonomy violation than addiction per se. Then, I argue that AE is

pervasive, practically necessary, and hidden, and that it objectionably interferes with people’s

second-order preference satisfaction as well as their internal preference formation. Moreover, I

argue that AE raises an obligational dilemma, namely that our obligation to attend properly to the

social and political world is at odds with our obligation to protect ourselves from the moral

harms of AE.

In Chapter Four (“Corruption and the Attention Economy”), I argue that AE expresses

and promotes corrupted attitudes about our attention and its value: particularly, that it is

instrumentally valuable, commensurable with market goods, and valuable solely for its quantity.

I argue for a threshold view of when this type of corruption occurs for certain goods like

attention—the pervasiveness and scope of AE plays a key role here. Importantly, in the process, I

defend my view from Brennan & Jaworski’s prominent critique of the corruption objection:

namely, that the meaning of markets is contingent and is of little moral weight.

Now, there is a theme that unifies these four chapters and that guides my research: What

underlies the moral problems with the attention economy? The recent expansion of markets and

market-like practices into central areas of our social lives depends upon and engenders the

economic way of thinking (EWOT), which I characterize as outcome optimization, the promotion

of narrow self-interest, transactional attitudes, and the commensuration of goods.

By delineating these features and explaining how this way of thinking degrades key

social practices (e.g., dating) and our ability to flourish, my work helps to precisify, expand, and

illuminate the corruption objection to markets defended by Sandel. Importantly, my work shows

how the crux of the corruption objection is about the ways people think, choose, and value—its

scope is not limited to practices involving the exchange of money.

III. Remarks

Before turning to the main chapters, I will make some general remarks regarding the

scope of this project, some caveats, its potential contribution, and directions for future research.

First, due to scope constraints, there are other morally problematic features of AE that are

not addressed in detail. These include algorithmic injustice, data privacy, and misinformation,

which are discussed in the ethics of technology literature, as well as harms to society such as

“polarization, balkanization, and extremism” (see C&P, 2020: 5-6). These are conceptually

distinct issues, and they are worries in themselves. However, they bear on our ability to live well.

It would be interesting to explore the connection between some of these issues and human

flourishing. Either way, to the extent that these other critiques implicate AE in impeding our

ability to live well, this further bolsters my moral case against AE.

Second, while I examine potential public policy interventions for AE, especially in my

chapter on dating apps, it is not within this project’s scope to analyze in detail whether there is a

better feasible alternative than AE, nor whether the social or economic costs of transitioning

away from AE would be too high. Given its serious moral downsides, I hope we are not stuck

with AE. However, it would be important to address these questions in order to best implement a

public policy response.

Third, my argument is not meant to imply that there are no benefits to AE and its

associated technologies and services. Relatedly, there may be unproblematic instances of AE

use, such as an isolated individual who is aided by social media in reconnecting with their

friends and family. My case against AE does not depend on an absence of benefits or that every

instance of AE use is necessarily wrong. Instead, what I aim to show is that AE’s cumulative

impact is, all things considered, significantly morally problematic.

Next, I will highlight how I hope this dissertation may make a contribution to the field. I

believe that this research project has the potential to add value along multiple dimensions.

Importantly, all of the core chapters examine underexplored topics in the discipline.

It is only in the past few years that there has been any serious analysis within philosophy

of the moral dimensions of dating apps. The literature on this topic is still in a nascent stage, and

my chapter “Dating Apps and the Limits of Markets” provides a novel take on it. Others have

argued that dating apps are addictive, gamified, biased, and unjust. Instead, I critique dating apps

in terms of their market structure and the degraded economic mindset that they promote, while

providing an analysis of why this technology and practice were adopted and became entrenched.

Also, while Castro and Pham’s pivotal article “Is the Attention Economy Noxious?”

(2020) raises important criticisms of AE, it mainly locates the moral problems at the level of

mental health, addiction, and exploitation. (It also argues for societal harms, such as

polarization.) Instead, I examine AE primarily from a virtue ethics perspective; I provide a

distinct moral critique by arguing that AE is at odds with virtues required for flourishing and that

AE expresses and promotes morally objectionable attitudes about attention.

In addition, my moral critique of AE and dating apps (an illustrative AE service)

enhances a theoretical understanding of the corruption objection to markets. Sandel (2012)

provided an influential account of this objection, along with a plethora of potential cases, from

people paying to skip queues to corporations naming public spaces. For some important goods

and practices, Sandel argues, putting them on the market would “express and promote the wrong

way of valuing them” (2012: 10). However, his account is not entirely clear about what this

corruption is and how exactly markets can corrupt. By explicating what EWOT is and how and

why it may corrupt important social practices, I endeavor to both illuminate the corruption

objection to markets and to make it more precise.

Further, I incorporate the interdisciplinary toolkit of philosophy, politics, and economics

(PPE) to shed light on some interesting questions. Why, for instance, can the uncoerced adoption

of new technology lead to an undesirable status quo and collective action problem? I use PPE

tools to enhance our understanding of why, for AE services like dating apps and social media,

technology promised one thing but delivered another.

Finally, my work offers a methodology for how to ask and address moral questions that

arise when important social practices intersect with digital media and EWOT. Both AE, in

general, and dating apps, in particular, are crucial cases of this intersection, and my analysis of

them incorporates and expands upon market critiques. This models how market critiques from

philosophy of economics can be applied to questions in the ethics of technology, which brings

together two subfields in philosophy. I believe that this methodology can then be expanded to

emerging technologies (e.g., AI) that promise to significantly impact our social lives.

To conclude this section, with an eye to future research, I will now raise a worry about

emerging technologies, such as AI and virtual reality. Namely, if we follow an AE model, which

uses attention as a central commodity, resource, and currency, then there is a serious concern that

the moral problems I argue for here will get even worse. This is because these powerful new

technologies risk supercharging the ways in which our attention is captured and diverted.

For instance, given its enhanced capabilities, AI has the potential to learn about us and

shape our preferences to a much greater degree than previous technologies. This risk will be

enhanced if AE is the model adhered to by the companies who design and profit from this

technology. For instance, AI chatbots, especially those designed to simulate close human

relationships, might be an even more powerful way to capture our attention. (See the movie Her

for a depiction of such a future.) We should be wary about companies having a strong profit

motivation to capture our attention via these technologies.

Although this needs further exploration, I hope to expand upon these ideas about AI and

virtual reality in future work. However, my dissertation already points to some key implications

for these new technologies. If I am on the right track, there is a strong case to be made that AI

and virtual reality should not follow the current AE model. At the very least, it would need to be

appreciably modified or regulated, perhaps by a mixture of public and private entities.

And it would be better to realize this early on, before AE’s model, as it has elsewhere,

becomes fully entrenched in this important domain. Because of the difficulty of collectively

changing the status quo once a technology and its economic model become entrenched,

preventative action may be crucial. An ounce of prevention may indeed be worth more than a

pound of cure. This is because, as I endeavor to show in the following chapters, AE is a

significant threat to key human values, including flourishing, intimate relationships, autonomy,

and properly viewing attention as intrinsically valuable and central to who we are.

REFERENCES

Bhargava, Vikram, and Manuel Velasquez. 2021. “Ethics of the Attention Economy: The

Problem of Social Media Addiction”. Business Ethics Quarterly, 31(3): 321-359.

Castro, Clinton and Adam Pham. 2020. “Is the Attention Economy Noxious?” Philosophers’

Imprint, 20, no. 17: 1-13.

Haidt, Jonathan. 2024. The Anxious Generation: How the Great Rewiring of Childhood is

Causing an Epidemic of Mental Illness. Penguin Press.

Newport, Cal. 2019. Digital Minimalism: Choosing a Focused Life in a Noisy World. Portfolio.

Nozick, Robert. 1974. Anarchy, State, and Utopia. Basic Books.

Pew Research. 2024. “Mobile Fact Sheet”. https://www.pewresearch.org/internet/factsheet/mobile/

Sandel, Michael. 2012. What Money Can’t Buy: The Moral Limits of Markets. Farrar, Straus, and

Giroux.

CHAPTER ONE: DATING APPS AND THE LIMITS OF MARKETS

I. Introduction

Dating apps have become the dominant way in the U.S. in which aspiring romantic

partners meet and attempt to form connections.1 Instead of meeting via (say) mutual friends,

people are using their smartphones to rapidly sort through hundreds of romantic options, each

comprised of a few pictures and a brief personal bio. If both users “swipe right” on each other’s

profile, they are allowed to initiate an in-app text conversation, which may result in a future date.

The potential upsides of this substantial social change may be easier to spot. By offering

something akin to a marketplace, dating apps significantly increase the number of romantic

options available to a person and reduce the difficulty of searching for new options. This increase

in dating options may help those who might otherwise have trouble meeting people. Also,

making it easier to leave a relationship may help some people find the confidence to abandon an

unhealthy or irreparable one. By creating a separate social sphere for dating, these features may

also help people avoid messy situations, such as workplace romances, and perhaps even reduce

unwanted romantic approaches in everyday life. Along related lines, Kugelberg (2025: 9) argues

that the design of dating apps “[has] furthered people’s interests in noninterference and choice

1 Rosenfield et. al (2019) found that, beginning in 2013, “meeting online has displaced friends as the main way

heterosexual couples in the United States meet”. Moreover, a higher percentage of “lesbian, gay, or bisexual (LGB)”

Americans have used a dating app compared to non-LGB ones (51 to 28 %). (Pew Research, 2023)

improvement” in the sexual sphere. Since dating apps seem to promise increased freedom and

choice in our romantic lives, one might think that they are generally beneficial.

And yet, there is a widespread sense that dating apps have serious downsides. Pew

Research (2019) found that dating app users are more "frustrated (45%) than hopeful (28%)”.

There is also something morally off-putting about these apps—a significant ick factor. For one

thing, online daters tend to think of themselves and others as market products. (See Heino et. al,

2010.) Moreover, a substantial number of dating app users complain of deceptive behavior and

superficial interactions. (Pew Research, 2019)

The general dissatisfaction with dating apps helps explain why Hinge—a popular app—

markets itself as “the dating app designed to be deleted”.

2 There is a strong desire for apps that

enable daters to find and maintain a committed relationship. Yet, many appear stuck in an

endless cycle of swiping, matching, frustration, giving up, and starting again.

But what exactly is morally problematic about dating apps? Some argue that the problem

concerns fairness and socio-economic justice: for example, Nader (2020) argues that dating app

algorithms reinforce bias, and de Vries (2023) argues that the apps’ “visibility boost” options

provide an unjust advantage to those with more means. However, here I incorporate market

critiques to argue that the structure of dating apps fosters inapt and degraded ways of valuing.

Specifically, I will argue that dating apps, by fostering a market mentality, undermine ways of

thinking that are needed for long-term commitments and properly valuing others in intimate

social contexts.

First, I argue that dating apps are at odds with the important goal of forming and

maintaining a committed relationship. I explain how this downside of apps (ironically) flows

2 See https://hinge.co/

from the same market features—increased options and ease of exit—that lead to some of their

apparent upsides. I also show how the attention-based business model of dating apps, which

relies on constant engagement, plays a key role in undermining commitment. Second, I explain

why these apps encourage users to think of themselves and others as market products. I argue

that this objectifying, transactional stance expresses and promotes degraded attitudes in this

intimate social context. My analysis elucidates how these two downsides are deeply connected:

underlying them is an economic way of thinking—which dating apps rely upon and encourage—

that is at odds with committed relationships and proper valuation.

Now, if these downsides are so bad, why do people remain on these apps? To answer this

question, I analyze systemic issues with dating apps, including how the current status quo arose,

why it may be hard to fix, and what might be done to mitigate these downsides.

II. A Mindset at Odds with Committed Relationships

A reasonable assumption is that many dating app users are looking for a committed

romantic relationship. Such a relationship promises to satisfy a core desire for intimacy, and

many consider it to be part of a good life. In fact, 44% of current or recent users reported using

dating apps “to find a long-term partner”. (Pew Research, 2023) However, only 10% of

partnered U.S. adults met their current partner on an app (2023), and “approximately 40% [of

users] ...reported being in a serious relationship while using [an app]”. (Alexopoulos et al., 2020)

If my initial assumption is correct, then what might explain these commitment issues?

Others, such as de Vries (2024: 3-5), have argued that the hiddenness of the companies’

algorithms, combined with their incentive to increase app use, makes it likely that their

algorithms are designed to prolong the dating app experience. While this hypothesis may have

some plausibility, I will show that it is unnecessary to speculate about the algorithm to get at the

core of the problem. Instead, the main reasons for these commitment issues stem from features of

dating apps that we can directly analyze, such as how their choice structure impacts daters. In

what follows, I will argue that the new choice framework and overall structure of dating apps,

reinforced by their attention-based business model, fosters a market mindset that is at odds with

forming and maintaining a committed relationship.

A) Options Overload and Exit vs. Voice

Crucially, dating apps change most users’ assessment of their next best alternative by

significantly increasing the number of potential romantic partners. The potential benefits of such

a change are clear—for example, having more options might lead to a better chance of meeting a

suitable partner. However, there are significant downsides to this new choice framework.

For instance, a user has more incentive to move on to a different person at the slightest

indication that their potential partner has an undesirable trait when there are so many other

options. Given that another option is so easily accessible, the likelihood increases that a dater

will lose interest in their date as soon as something seems wrong. But this choice pattern, while

seemingly rational in the short run, may lead to long run results that are at odds with the desired

result of a committed relationship. A deep romantic connection may be likelier to occur when

both parties are only dating one person and over a longer period than dating apps encourage.

Underlying this thought is that a certain amount of psychological commitment and proper

attention is generally a prerequisite for a meaningful, lasting romantic connection to occur. And

to some extent, it needs to happen throughout the dating process, including the choosing stage,

not merely after the relationship becomes serious. So, the large increase in options may instead

be a recipe for users remaining on the apps for a long time, going from person to person, without

ever fully committing to getting to know any one person.

Now, one might say that love has always involved a time for choice followed by a time

when the choice has been settled. While this is true, the choice structure of dating apps is such

that it makes the attitude of continually looking for options harder to drop. If the early stages of

love involve so much choice, it becomes harder to switch to the type of mindset required for

commitment. Part of the worry here is that the culture of dating apps conditions daters into

thinking that their love lives are a terrain filled with perpetual choice.

This concern aligns with Schwartz’s (2016) argument that having too many options can

be paralyzing, given our psychological tendencies. Regarding our love lives, Schwartz (2016:

39) posits that an increase in options could make it harder to choose one person because you may

miss out on someone “better”. This tendency to avoid regret supports my claim that the

commitment relationships require is in tension with dating apps’ vast array of romantic options.

Moreover, dating apps add to the tension between commitment and an abundance of

choices by making it relatively frictionless to move on to someone else. They significantly

reduce the time and energy spent searching for a new partner, because you only need to pull out

your phone and start swiping. That is, apps make it easier to leave one’s current dating situation.

Following Hirschman (1970), there are two main ways of dealing with dissatisfaction:

people can simply exit a situation, or they can stay and use their voice to communicate with those

involved. Hirschman claims that people will use voice when exit is difficult and when they think

that voice will work. Importantly, exiting when dissatisfied is characteristic of market

interactions (e.g., changing restaurants without lodging a complaint) whereas voice is

characteristic of social relationships (e.g., expressing our feelings to a loved one).

As noted, dating apps make exit much easier. This enhanced freedom to exit explains

further why dating apps are at odds with committed romantic relationships. The concern is that

dating apps encourage a shift from a norm of voice to a norm of exit, whereas being willing to

prioritize communication over exit when dissatisfied is part of the process of forming and

maintaining a romantic relationship. There is a similar tension between commitment and

freedom in other domains. For example, economic mobility may lead people to be less

committed to their communities. By analogy, increased “mobility” (ease of exit) on dating apps

may weaken a more committed way of relating to a romantic partner.

In sum, because dating apps make exit easier and add many more options for potential

partners, they weaken the motivation to give any potential partner a chance. Now, there may be

positive aspects to this new choice structure, such as providing more options to isolated

individuals. But the increase in options and ease of exit incentivize people to give up on a

relationship that could be salvaged. Relatedly, it discourages people from devoting enough time

and attention to one person in a new relationship to form an enduring, committed relationship.

The new choice landscape of dating apps may result in a society with more perpetually lonely or

dissatisfied people.

Several representative studies from social psychology support this explanation of why

apps are at odds with commitment. For instance, D’Angelo and Toma (2017) found that online

dating results in choice overload—the dissatisfaction of daters increases with more choice.

Participants were either shown six or 24 potential partners. Those shown 24 options were not as

satisfied and less likely to adhere to their initial choice.3 This effect was further enhanced by

3 Also, Pronk and Denissen (2020) found evidence of a rejection mindset—as an individual was presented with a

sequence of random dating app profiles, they were less satisfied with and more likely to reject each additional

option. This suggests that actual app users, who are continuously shown new options, will develop this mindset.

reversibility—participants who were given an opportunity to change their choice later were even

less satisfied. This result applies to dating apps in general, because app users typically have a

very large set of options and can always reverse their initial choice by moving on to the next

option. Given these two features, the study suggests that dating app users will experience a high

level of dissatisfaction with whomever they are currently dating. If so, this will make it harder

for them to substantially commit to that person.

B) An Optimizing Mindset

There is a related way in which dating apps interfere with forming committed

relationships. Due to the choice structure of dating apps, interactions between potential partners

are significantly more calculated and planned. They have deliberately chosen to meet, based on

thinking through the desirability of the other person’s advertised qualities. However, this

optimizing approach may backfire. Although optimizing seems more efficient, organic

relationship development may play a key role in creating and sustaining romantic connections.

The suggestion I would like to make is that an initial “spark” or romantic connection may

be more likely to occur when it is not being actively sought out. This is like the paradox of

hedonism: the harder one tries to be happy, the more elusive happiness becomes. When romantic

connections are sought in an overly deliberate manner something similar may occur. The app

framework and choice environment encourage a mindset that is at odds with approaching dating

in an organic, less calculating way. Ironically, approaching romantic relationships more

“efficiently”—via a calculating, optimizing mindset—may actually turn out to be less efficient.

Now, one might object that finding romance via organic development—like slowly

getting to know someone informally via a shared activity or friends—might have its own flaws,

and that approaching dating in a more calculated way could address some of these flaws. For

instance, it could be argued that basing romance on organic connections, without sufficient

advanced consideration of compatibility, such as values or personality type, is a recipe for failed

relationships. The worry is that a person might end up with someone who is not a good long-term

fit for them if they look for love in a less calculating way.

In response, many personal details that dating apps make tangible and salient are not

necessarily indicative of a good romantic match. For instance, searching for partners with a

desired physical appearance, education-level, or hobbies, might not help much in predicting

romantic fit and relationship longevity. These details do not provide enough information for this

purpose. If someone relies too heavily on them, they risk screening off good potential

connections.

But perhaps there are other features, such as core aspects of someone’s personality or

their serious interests and values, that are more helpful indicators of future relationship success.

And there is something to be said for the notion of balance—that, even in romance, one must

listen to one’s head in addition to one’s heart.

In response, voice is needed to ‘get at’ these deeper, more intangible indicators. One must

commit oneself to having meaningful conversations with the other person. This is not only

because meaningful conversations reveal these deeper features, but also because they are needed

to work through the setbacks that arise when building a committed relationship. However,

because apps encourage exit over voice, it is harder for people to take the time and energy to

determine these important features and to work through difficulties as they arise.

Moreover, romantic connection is not a fully rational process. One cannot determine

compatibility in advance, based on any set of salient features. Romantic connection is an

intangible good that is co-created by both parties. It is different than a good whose nature and

value can be determined in advance, such as many material goods, and whose value is due to

readily identifiable features. If so, this helps explain why it is hard to know whether you will

connect with someone before meeting them in person, and why people who seem “good on

paper” often fail to connect. In a good romantic match, there seems to be something both

intangible and holistically developed over time. These features cannot be determined beforehand,

and many can only be realized by using voice to work through setbacks. An optimizing approach

to forming romantic connections fails to capture this aspect of enduring relationships.

At this point, there is an important distinction to be drawn between two issues raised by

other authors and my above analysis. Klincewicz et. al (2022: 557) argue that dating apps are

designed in a casino-like way to “facilitate behavioral addiction”. Relatedly, Nader (2024) argues

that dating apps have “gamified” dating—that is, that the design of these apps changes some

users’ goals from forming a romantic connection to simply accumulating the most matches.

While these issues—to the extent that they are occurring—could exacerbate the commitment

problem I have raised, they do not get to its root. This is not to say that these issues aren’t an

additional concern. In fact, I turn next to how the financial incentives of app companies play a

contributing role in inhibiting commitment. That said, even if users were neither addicted nor

using the apps for the purpose of simply collecting matches, the structure of dating apps—by

encouraging exit and optimization—would still significantly interfere with commitment by

fostering this market mindset.

C) Contributing Problem: The Attention Economy

My analysis of how dating apps inhibit commitment leads naturally into examining the

attention economy4 business model of online dating. Dating apps are maintained and promoted

by for-profit corporations, such as Match Group. Regarding the commitment problems that more

options and easier exit raise, these “free” dating app companies are more than happy to capitalize

on our inability to commit to one person. Their primary aims, due to their attention-based profit

model, are to keep users on the site so that they can harvest their data to sell to advertisers and to

entice users to sign up for a paid subscription.

Dating apps are an example of what goes wrong when the social purpose of an activity is

in tension with the profit motive of the company providing a related service. A driving

motivation of dating for many people is to form a committed relationship. However, the driving

motivation of dating apps, under the attention economy model, is to keep users on them for as

long as possible. This is a problem of perverse incentives, where the primary incentives of the

company may generate undesirable results for consumers: here, the apps perversely want users to

continue dating for as long as possible, while many users aim to find love and get off the apps.

To take a key example, Hinge’s advertising slogan is misleadingly titled “designed to be

deleted”. A reasonable interpretation of this is that Hinge’s main aim is to help users find love

and quickly delete the app. This slogan is disingenuous; it hides their main motive, which is to

monetize their users’ attention via constant engagement, which in turn requires users to stay on

their app and not delete it.

4 I roughly follow Castro and Pham’s (2020: 2) description of the attention economy: a market where [1] people

exchange their attention for access to a new media service (e.g., Instagram); [2] these companies then sell our

attention, and the data it generates, to advertisers.

Consider how these apps continuously show users new potential matches, regardless of

how many matches they already have. This constantly prompts the user to think about whether

there is someone else who is better suited for them. This attention-grabbing feature relies on

weaknesses in human psychology, such as the “fear of missing out”. (cf. Castro and Pham: 2.) It

makes it harder to commit to one person when the app constantly reminds you that there may be

a better option out there if you only stay on the app. If an app company’s purpose was to help

users delete them quickly, then it should substantially limit the number of people one can be

shown per day or match with at one time. However, Hinge and most other apps do not do this,

and this is no accident. Instead, their incentives help explain the attention-grabbing features of

dating apps, which contribute to a mindset that is at odds with committed relationships.

One might object that, for people to keep using them, the apps must give people what

they want, including committed relationships. To be fair, the apps must sometimes work in this

way for people to continue using them. I propose, however, that apps require user success only to

the extent that this will allow them to successfully hook the most people to their app for the

longest amount of time. In fact, dating apps only need to appear to work, and it is hard for users

to know how well they do. Relatedly, given the large number of users, people are bound to know

some dating app success stories, and apps seem to show users a skewed ratio of highly desirable

potential matches. This may encourage users to overestimate their likelihood of success.

Moreover, the possibility of romantic success, even if unlikely, is enough to attract many

users, given how central the human need is for love and connection. This is especially true given

how widespread the use of dating apps is and how it has diminished other avenues for meeting

people. Someone looking for love may feel like they have few other feasible options but dating

apps, because apps are now the cultural norm. Daters are in a poor bargaining position in that

they largely have to use the apps to satisfy a basic human need.

Dating apps hook people’s attention by how they structure their options. Take another

feature of dating apps: it appears that none of them allow the user to post an away message. The

purpose of an away message is to alert other users that they shouldn’t expect a fast response.

Without such a feature, the user is not able to free themselves from the expectation that they

should respond in a socially acceptable amount of time. Given our cultural assumption that

people always have their cell phones on them, the amount of time in which they are supposed to

respond is limited. So, users will be strongly induced to constantly check their apps and respond

to any messages. This is a recipe for the continuous engagement of our attention. It highlights

how dating apps work on the attention economy model.

The structure of messaging is one of many processes that habituates people to constantly

attend to their smartphones; the psychology of individuals is affected by the framework of the

attention economy. This is a vicious feedback loop of desire formation and solidification—the

desire to check the phone is rewarded by achieving some good (e.g., a potential connection),

which enhances the desire, and so forth. Over time, the behavior of checking these apps becomes

a habit or even an addiction. Importantly, the profit motive of the attention economy model

incentivizes companies not to institute mitigating features (e.g., an away message option)—

because this would inhibit habit-forming processes.

To recap: the choice structure of dating apps, reinforced by the attention economy, fosters

a habitual market mindset of exit and optimization. The key downside is that this mindset is in

deep tension with the goal of a committed relationship.

III. Problematic Norm Changes: People-Shopping

Here, I will expand upon how dating apps have turned dating into something more akin to

regular shopping, with a focus on how this corrupts our attitudes and character. Given how much

scrolling, swiping, split-second judgments, etc. apps encourage, the users’ valuation of others

may shift due to this market ethos. Online dating is like people-shopping—on apps such as

Tinder and Hinge, each person is a little picture on a phone who is evaluated in a very short

amount of time given a limited amount of surface information. This is primarily the other user’s

pictures, along with some basic details regarding their interests, occupation, etc. The problem is

that this rapid-fire sorting of other people based on limited information—without in-person

communication—may distort people’s attitudes towards others. For instance, it may encourage

users to think of other people as a bundle of traits that may or may not satisfy the users’ criteria.

Importantly, this economic way of thinking makes online dating a market in most respects, even

though no money is exchanged between users.

Empirical studies provide support that dating app users exhibit this people-shopping way

of thinking. In a key in-depth qualitative study of online daters, Heino et. al (2010) found that

market-like attitudes and strategies are common. In particular, the authors found that daters

frequently: 1) interpreted other profiles as a resume or advertisement; 2) quantified items

indicating their own market desirability, such as number of matches or messages; 3) focused on

tangible and discrete features (e.g. age, job, etc.); 4) thought of dating as a numbers game—they

kept trying people to optimize their result, and often dated multiple people at once. Notably,

daters thought of themselves as a part of the dating market.

Of course, it is plausible that people were doing this to some extent before the rise of

dating apps. The problem is that dating apps significantly exacerbated this feature of dating,

making it central, while minimizing the features—such as in-person communication and social

norms—that mitigated the tendency to objectify and instrumentalize people. Consider how

dating apps bear a striking resemblance to other versions of online shopping (think of it as

“Amazon for people”). In many ways, including how they display queues for potential and

existing matches (“options”), they look like just another variation of these shopping apps.

Relatedly, the dating app context minimizes an important check on instrumentalizing,

market-like behavior. When people meet via friends, or via some other social group (religious,

family, neighborhood, etc.), there are built-in constraints because the daters share non-trivial

social connections in common. This provides reason to treat each other with a minimal amount

of respect, instead of merely as an option to be sorted through. If people meet via mutual friends,

they have an external reason to avoid tactless behavior—presumably, they won’t want their

mutual friends to think less of them. There is more incentive to (say) have a potentially difficult

conversation instead of abruptly ending all communication.

But with dating apps, people regularly match with others with whom they have limited or

no outside connections. Mutual social connections are not built into the dating app context, as

they were with most previous dating practices (e.g., meeting via friends). Removing shared

social connections from the dating context loosens some of the social norms involved. As with

dating apps’ online shopping-based features, taking away this social embeddedness contributes

to the sense that dating is more of a market than a different type of social practice.

This connects to my point about exit versus voice. Dating apps facilitate a market

mentality where users feel free to immediately reject a potential partner without much thought or

care. This is both before and after meeting in person. If so, dating app users are following the

norm of exit in the face of non-optimal choices. This norm encourages people to think of others

as options to be sorted through and discarded if not optimal.

Further evidence that many users of dating apps have experienced this shift in their

attitudes towards others includes the common phenomenon of “ghosting”: where one dater stops

communicating with another without any warning or explanation, even after meeting multiple

times. A change in attitudes towards others, in an instrumental direction, helps explain why

people feel like this is an acceptable way to behave. In this way, ghosting is a symptom of the

underlying cause—a people-shopping mentality.

A) Dating Apps and the Corruption Objection to Markets

By fostering a people-shopping mindset, the market-based features of dating apps shifted

individual attitudes and dating practices; and I will argue that these shifts degraded them. This

position is in line with what Sandel (2012: 9) calls the corruption objection to markets. Sandel

argues that it would be morally wrong to (say) buy and sell children for adoption in a standard

market, even if there were no coercion, inequality, or mistreatment. Such a market would

“express and promote the wrong way of valuing [children]”; they are properly valued as persons

“worthy of love and care”, so valuing them in a profit-focused way would be wrong. (2012: 10)

There are two argumentative steps needed to establish that the market aspects of dating

apps are corrupting. The first is that, as argued above, the choice structure and framework of

dating apps affect people’s mindsets—that the widespread adoption of apps was not causally

neutral with respect to individual attitudes and social practices. Now one can debate whether this

is a good or bad thing. But to admit that these apps have shifted attitudes and norms is to admit

that they are not merely reflecting but instead driving preferences, values, and culture. Sandel

(2012: 125-126) claims that one tenet of market faith is that markets are innocuous, that they do

not shape or change who we are, our values, or our social norms. But, as I have argued, dating

via apps (a market-like practice) does shape its users in these ways.

The second prong of the argument is that these shifts in preferences, attitudes, norms, and

practices are morally degrading or diminishing. Now, Klincewicz et. al briefly consider the

possibility that dating apps have corrupted people via their “mercantile…logic”, before

suggesting that apps have merely dispelled the “illusion of fate” that was veiling the reality that

dating is like “a marketplace casino”. (2022: 567-68) This analysis is inadequate—there is much

more to the corruption objection to dating apps. Regardless of whether a belief in romantic fate is

valuable, I will argue that important values have been lost or diminished by the mindset shift

engendered by dating apps.

My main claim is that dating apps, by promoting a market mentality (“people shopping”),

undermine ways of thinking that are needed for long-term commitments and properly valuing

others in intimate social contexts. To better support this claim, it will be helpful to elucidate and

expand upon Sandel’s corruption objection to markets. So, I’ll now state precisely the core of

what I think this market mentality is—I call it the economic way of thinking (EWOT). It is an

approach to choosing and valuing characterized by a) outcome optimization; b) narrow selfinterest; c) transactional attitudes; and d) the commensuration of goods.

Dating apps foster EWOT. We’ve already seen how they encourage users to optimize

their dating outcomes as well as to instrumentalize other people, which expresses and reflects

narrow self-interest and a transactional attitude. I now argue that by expressing and promoting

EWOT, dating apps degrade people’s attitudes and values in the intimate context of dating.

First, there are deep connections between the people-shopping attitude of dating app

culture and the problem of commitment. Importantly, thinking of potential partners as

commensurable or fungible (replaceable), which is part of people-shopping (e.g., continuously

comparing people via tangible features), helps to explain the commitment issues. This type of

thinking is at odds with the kind needed for commitment. Instead, commitment requires that one

thinks of one’s partner and relationship as irreplaceable. But this is harder to do when

relationships are formed in the context and environment of people-shopping. This market-like

practice requires participants to constantly evaluate and choose between goods (i.e., a

prospective partner), which encourages them to focus on and value commensurable aspects and

to discount any non-commensurable ones. Relatedly, encouraging rapid comparisons via a

handful of salient features also diminishes a holistic appraisal of any potential romantic partner.

EWOT in dating involves optimizing one’s romantic options by comparing their

attractive features, which is an inapt foundation for forming and maintaining committed

relationships. We may value a partner’s attractive qualities; but if we only value them for these

qualities, then we don’t love them appropriately. Focusing primarily on a partner’s desirable

features is an inconstant and unstable approach to relationships, because these features (even

deeper ones) are subject to change. Particularly, when EWOT dominates, people are more likely

to trade up for a “better” partner (more attractive, etc.). But this would be the wrong way to value

romantic partners. In any plausible model of love, commitment involves thinking of a partner as

not replaceable in this way. To form an enduring bond, one needs to be willing to substantially

commit to someone. One’s commitment may be revisited, but neither lightly nor often. Dating

apps inhibit cultivating this important way of thinking, a mindset essential to forming the

committed romantic relationships that are part of a flourishing life for many people.

Second, and relatedly, by fostering EWOT, dating apps encourage inapt thinking about

the nature and value of the romantic relationship itself. These relationships are an important

example of a general distinction between goods—individualistic versus shared. Roughly, I take

shared goods to be those that don’t belong to any individual but to the social group; in turn, they

can only be fully enjoyed by the group together. In this case, romantic connections and

relationships are built together; they don’t belong to either person. Moreover, the good can’t be

enjoyed in its fullest form unless the couple is enjoying it together.5

Shared goods are often constructed by groups of people—they are brought into existence

by mutual collaboration between individuals. In the case of relationships, they exist in a shared

social “space” between the individuals. Lasting romantic connections and relationships are cocreated by the individuals over time.

This helps to explain why the narrowly self-interested, optimizing EWOT (i.e., peopleshopping) approach encouraged by dating apps may not work very well. Shopping in advance for

a relationship and selecting people to meet doesn’t yield any of the desired good, because a

connection and a relationship come into existence through the shared creation of two individuals.

In this way, it is a different type of good than, say, a food item, which can be easily purchased at

a grocery store. Food items can be treated as individual goods, as well as private property, that,

once obtained, don’t require another person to share in their creation and enjoyment.

One potential objection is that romantic relationships are reducible to a bundle of

individual goods that two people trade, and that to think otherwise is to obscure the reality of

such relationships. (cf. Becker, 1976.) This is a radically different conception. What does it

miss? For one thing, the strong sense in many relationships that one’s well-being is inextricably

5 For a related account, see Anderson’s (1993) discussion of types of goods in Values in Ethics and Economics.

tied to one’s partner’s, and that the relationship itself can be doing well, suggests that a robust

romantic relationship is a shared good that cannot be reduced to its parts. For another, the

existence and health of a committed relationship is bolstered by viewing it as neither

commensurable with nor replaceable by another one, and this way of thinking is in tension with a

transactional, preference satisfaction approach to relationships.

Third, from a virtue ethics perspective, romantic relationships are about helping each

other grow and become better people. Given their relative centrality in many people’s lives,

romantic partnerships are supposed to be deep and important relationships. These types of

relationships should involve challenging and shaping one another—including one’s preferences

and character—not just satisfying pre-existing preferences. This aligns with how Aristotle

conceived of complete friends—friends we admire for their virtue as opposed to friends who

merely bring us pleasure or utility. One key element of virtuous relationships is that we admire

our partners and grow along with them; the purpose of these important relationships, including

flourishing romantic ones, goes beyond the satisfaction derived from our partner.

The contrasting view that dating apps promote about relationships is that they are

primarily about optimizing individual preference satisfaction. Admittedly, it can be debated

whether the proper purpose of romantic relationships is about more than this. However, dating

apps rely upon and promote this self-interested and transactional EWOT attitude about

relationships without defending it. To the extent that dating apps shape our character and

attitudes, we ought to demand some accounting for why it’s okay to think about romantic

relationships in this market-like way.

A fourth consideration is that, in part due to its lack of social embeddedness, this new

dating context relies upon and promotes an atomistic, individualistic way of viewing oneself and

others (i.e., the narrowly self-interested and transactional aspect of EWOT). But it is

questionable whether this conception is the best one to hold in this domain of social life, given

dating’s close connection to intimate relationships, as well as to community life. The concern is

that EWOT, when inappropriately or excessively adopted by individuals, may degrade our

communal life. Individual choices, at least when it comes to dating, cannot be isolated from the

larger social environment. Attempting to separate the two, via the market attitudes promoted by

apps, can lead to problematic cultural changes. For instance, expanded markets and EWOT in

dating can disrupt cultures by further releasing people from previous social obligations. To the

extent that a valuable common good—such as a closely-knit and supportive community—is

diminished via these changes, this is another reason why treating too many social practices (such

as dating) in market-like ways may degrade our communal life. Removing the social

embeddedness of dating may further exacerbate the loss of community that has already been

happening in many other important areas of our social lives.

In this section, I expanded upon how dating apps rely upon, reflect, and encourage

EWOT, and I argued that this degrades important attitudes and values central to dating and

romantic relationships. Importantly, I also provided a key illustration of how the corruption

worry runs deeper than markets that involve the exchange of money. Relationship seekers on

dating apps are encouraged to think of themselves and others as a marketable product. While the

economic way of thinking may be unobjectionable when one is (say) shopping for groceries, I

have argued that EWOT, which is induced by dating apps, is morally problematic for the social

practice of dating. Moreover, the case of dating apps highlights how the corruption objection to

markets is ultimately about the ways people choose and value. If I am correct about this, the

corruption objection is not limited to practices where money is exchanged.

IV. Dating Apps and Technology-Induced Change

Now, if the above downsides are so bad, why do people continue to use dating apps? One

might think that if people really wanted to date in a different way, they would. In response, I will

present a systemic account of why dating apps diminished other dating modalities and why

fixing these downsides may be difficult.

A) Upsides Initially More Apparent

A simple story about the adoption of a new technology is that it has provided a more

efficient way of doing things. People chose to switch to a new technology because the benefits

outweighed the costs of doing so. Moreover, those who did not make the switch suffered no loss,

because they could continue doing things the old way. This would be a Pareto improvement: no

one is worse off and at least one person is better off compared to the previous situation.

However, I argue that, in the case of dating apps, this simple story is not an accurate one.

As we’ve seen, dating apps do not deliver increased preference satisfaction to all daters. In fact,

dating apps are an important example of the general phenomenon that new technology often

promises one thing while delivering another. Their potential upsides, such as more options and

easier exit, were more apparent up front. And these were promoted by the dating app companies

banking on their product’s adoption. On the other hand, the downsides—such as a deep tension

with commitment and the norm of people-shopping—were not as obvious to users, in part

because these problematic trends did not take place immediately.

Now, when the upsides to a new technology (e.g., dating apps) are more apparent up

front, and the downsides hidden or unknown, a social group may fall into a problematic status

quo. Once these new practices and norms become widely adopted, it is hard for any individual

acting on their own to escape engaging in them, let alone change other people’s attitudes.

Importantly, this new status quo will be hard to change even if most or all individuals are

unhappy with it. That is, at least unless they have a relatively low-cost way to organize.

However, dating app users are a large, widely dispersed, and heterogenous group; there is no

users’ society or lobby. So, there would be substantial costs to organizing. On the other hand, the

app companies benefitting from the status quo are much smaller, more organized, and cohesive.

This situation is a collective action problem—an individual cannot unilaterally deviate

from the current way of dating and improve their own situation, even if there is another status

quo most daters prefer. Given this and the importance of a romantic relationship for many

people, it will be hard for an individual to act counter to the new way of doing things.

The story here is that dating apps seemed like a good idea—the upsides were more

apparent up front. This encouraged many people to start using the apps, and there was no initial

problem in terms of preference satisfaction. However, some people’s preferences for how to date

changed once the downsides become apparent. But then they were stuck in the new equilibrium

and could not unilaterally improve their situation or the culture of dating. So, what started off as

an apparent improvement led to a new equilibrium where some people are less satisfied than they

would have been.

How might we model this process? The introduction of dating apps created the following

dilemma. Adding the choice to date via apps allowed each person an additional dating mode

option, beyond the extant in-person methods. For each person, no matter what others were doing,

it was a better strategy to focus on the new app method instead of solely relying upon the extant

ones—this was the dominant strategy. After all, at the low cost of setting up a profile, apps gave

each user many more dating options than they previously had. It also enabled avoiding the

uncomfortableness and uncertainty of trying to meet people in everyday life.

But once everyone started dating that way, it became much harder to engage in the

previous way of dating; the culture and practices surrounding dating shifted. People lost the

option of engaging in other forms of dating to the same extent. In this respect, everyone is now

worse off than they would have been if no one had adopted the strategy of primarily dating

through apps.

B) Shifted Choice Sets

Dating apps shifted the feasible choices available to individuals for how to date. Because

of the rise of dating apps, it is less acceptable for people to romantically connect with one

another in “real” life. Now, it often is practically necessary to date via apps, regardless of one’s

preferences. So, “adding” the option to meet people on dating apps didn’t expand people’s

romantic option set. Instead, it shifted it—certain options like in-person courtship have been

diminished by prevalence of dating app culture.

Dating apps led to the adoption of new norms and workable dating strategies. Consider

someone who would prefer to date one person at a time. Dating apps have made it more likely

that whoever they are dating may also be dating many others. Apps encourage this strategy and

offer many potential partners, so the person that they are dating has less incentive to commit to

getting to know them in return. In this new environment, because many others are “playing the

field”, it becomes riskier to invest in dating one person at a time. Relatedly, some people may

prefer to date in a more organic, informal manner. But if few people are doing so, this forecloses

the feasibility of a more spontaneous approach towards discovering a romantic connection.

While this is not to say that previous dating practices were unproblematic6 (e.g., in-person

approaches can be invasive), these cultural changes make it harder for those interested in the

above forms of dating to do so.

Consider how letter writing has been replaced by email as the socially acceptable way to

send many messages. The benefits of such a change are clear: it is much more efficient.

However, there are downsides: people now find it strange to receive a letter from someone.

Email technology changed social norms, which shifted people’s options. It may seem like the

number of choices people have merely expanded, but on closer examination some options have

disappeared. Writing letters is no longer a feasible choice in many cases.

Analogously, while dating apps appear to merely add options, they have in fact shifted

the option set for many individuals from meeting people in-person to meeting them online,

because it is harder and less socially acceptable to meet potential romantic partners via in person

encounters. If so, those who would prefer to date this way are not better off.

This point generalizes: adding options to a social practice, especially via new

technologies, may diminish or crowd out previous ways of doing things. The simple story that

adding options makes everyone better off or the same falsely assumes that adding options

doesn’t inhibit previous ways of doing things. When adding options to a social domain shifts its

practices, some people are worse off.

Here is the more nuanced story about dating apps. Given the prevalent online dating

culture, one often must use dating apps and use them in certain ways, regardless of one’s own

preferences. Counter to a core assumption of economics, adding options sometimes results in

worse outcomes in terms of preference satisfaction. Some people may prefer to live in a world

6 See Kugelburg (2025) for further discussion of some problematic aspects of previous dating practices.

without dating apps, where other options for how to date are more available or salient; but given

the ubiquity of apps, they prefer to use them rather than not use them.

C) A Challenging Collective Action Problem

Importantly, app companies do not want to move away from the status quo. Even if many

customers would like to get off the apps, the attention-based business model of companies

incentivizes them not to change the features that hook users. Also, there are likely at least some

people who prefer the current equilibrium to any other feasible alternative. For example, perhaps

some people are fine with how dating apps are at odds with commitment.

Moreover, given the structure of the apps, some users may start to prefer novelty and

excitement over focusing on one individual at a time. Their preferences about how to date and

their conception of the value and purpose of dating may change. They may come to believe that

there are always more or better options available, and they may form shorter attention spans for

considering partners. This is because the attention-capturing features of dating apps encourage a

preference for continuous shopping for romantic partners, analogously to how different social

media apps encourage frantic scrolling for other objects of our attention. In the face of such

features and pressures, people might come to have adaptive preferences and attitudes. They

prefer the status quo, but only because their preferences changed via a process they would not

reflectively endorse or because they modified their preferences to avoid further disappointment.

That said, many people may still be happier if dating practices changed to mitigate the

above downsides. However, a smaller number would not be—i.e., those who are benefitting the

most, especially the app companies. But these companies are the actors who have the most

power to change the status quo, because they are more organized and more resourced than the

heterogenous group of daters. So, even if most people are unhappy with the equilibrium, given

the difficulty of organizing and the opposing interests of app companies—in addition to shifted

choice sets and adaptive preferences—app users face a challenging collective action problem.

I’d like to suggest that this is generalizable—that dating apps are just a key example of

technology promising one thing but delivering another. These systemic issues apply to other

attention economy services like Facebook and Twitter (X). In both cases, the apparent upfront

benefits (e.g., socializing and information) had associated long-term downsides (e.g.,

polarization and misinformation), due in part to their economic model, including the perverse

incentive to maximize user attention. Because of opposing interests, adaptive preferences, etc.,

changing the status quo for how we socialize and share information is now a challenging

collective action problem. These systemic worries provide reason for caution about other

technological intrusions (e.g., generative AI and VR) into important social practices like dating,

especially when driven by for-profit companies with interests unaligned with the common good.

D) Mitigating the Downsides

Here I will examine some suggestions for how to mitigate the downsides, starting with

what individuals could do. Dating app users could institute certain types of self-regulation. This

includes limiting the amount of time spent on apps and reducing the number of people one talks

to at a given time. These steps may help mitigate the ways in which the dating app environment

inhibits a person’s ability to commit to someone. By (say) limiting one’s time on these apps, the

ever-present pull of easily accessible alternatives might be lessened.

However, self-regulation wouldn’t help for certain downsides. For instance, one would

still be participating in the practice of people-shopping, even if one were trying to limit the

amount of time spent doing so. Moreover, given the systemic issues involved, unorganized

individual action would not significantly alter the culture engendered by dating apps. Resolving

these downsides seems to require a response at the social and public policy level.

The least invasive public policy response would be to educate and inform citizens of

these potential downsides. Also, a warning label could be added to dating apps outlining what

are taken to be the main downsides. Moreover, while it raises potential autonomy concerns, a

further suggestion would be for dating apps to incorporate some nudge features into the apps. To

the extent that it would be in their best interest (by their own lights), people could be nudged

away from falling into the commitment pitfalls discussed above.

Let me sketch some potential ways the choice structure of dating apps, such as Hinge,

could be changed to help mitigate this problem. First, they could institute a version of speed

dating, where users talk to multiple people in one hour, and then immediately decide whether

they’d like to speak off the app. This would eliminate much of the time and energy expended on

apps before meeting in person or over the phone. Second, apps could give people a limited

amount of time for their online connection to last—this would encourage people to get off the

app and meet elsewhere. Third, they could only allow someone to match with a small number of

people at once. This would encourage deeper connections over long-lasting scrolling.

Instead, as with other popular apps, Hinge’s framework involves many choices and easy

exit, and its features are structured to encourage user engagement via habit forming processes.

All of this is evidence that Hinge is not actually designed to be deleted; instead, it hooks its users

into paying constant attention to its services. It attracts people who would like to find committed

relationships, but then it does not institute features that would make this more likely.

We might ask whether a more substantial public policy intervention is warranted, given

the app companies’ perverse incentives and the collective action problem daters face. For

example, de Vries (2024) argues that introducing a well-regulated, state-run dating app

(alongside commercial ones) would mitigate the problem that apps prolong people’s romantic

searches. However, this proposal both goes too far and not far enough. It goes too far because

within Western liberal democracies allowing the state to directly manage its citizens’ intimate,

romantic lives, especially to this extent, is an undue violation of individual liberty. Yet, it does

not go far enough because, besides eliminating certain features like paid visibility boosts, the

state-run dating apps de Vries proposes appear to have a similar market structure (large option

sets, etc.) to the commercial ones we’ve examined. These apps would still engender peopleshopping, even if the state were administering them. These structural issues are the core of the

downsides examined in this paper, and a successful intervention would need to change them.

Moreover, there is a proposal that would address these structural issues while being less

invasive to individuals: namely, taxing or regulating some attention economy incentives,

including profiting from the time users spend on apps. This revenue could (say) be subject to an

additional tax, which would induce dating app companies to generate revenue in less attentioncapturing ways (e.g., a one-time enrollment or success fee). This tax would lessen the perverse

incentives problem and encourage companies to redesign their choice structure. I’m not claiming

that this is a practically feasible strategy; it would threaten the core of the attention economy

model, so there would likely be substantial political resistance from the companies affected.

What I am suggesting is that if such a policy were instituted, then this may mitigate the

downsides. I leave the feasibility issue for further consideration.

V. Conclusion

In this essay, I argued that the market-oriented features of dating apps have substantial

downsides for dating practices: namely, the obstruction of committed relationships and a culture

of people-shopping that expresses and promotes degraded attitudes. These downsides stem from

significant increases in potential romantic partners and ease of exit, and they are reinforced by an

attention economy business model. By encouraging the economic way of thinking, dating apps

undermine ways of thinking required for long-term commitments and properly valuing others in

intimate social relationships. Even if their algorithms gave users a fair chance, and users were

neither addicted nor treating the apps as a game, these core downsides would remain. Moreover,

I explained why these downsides are embedded within systemic issues, such as shifted choice

sets and a challenging collective action problem. I also sketched some ways of mitigating these

problems. I endeavored throughout to highlight why we need to develop a healthier relationship

with the social practice of dating.

REFERENCES

Alexopoulos, C, Timmermans, E, and McNallie, J. 2020. “Swiping More, Committing Less:

Unraveling the Link between Dating App Use and Intention to Commit Infidelity”. Computers in

Human Behavior 102: 172–180.

Anderson, Elizabeth. 1993. Values in Ethics and Economics. Harvard University Press.

Aristotle. Nicomachean Ethics. Translated by David Ross. 1998. Oxford University Press.

Becker, Gary. 1976. The Economic Approach to Human Behavior. University of Chicago Press.

Castro, Clinton and Pham, Adam. 2020. “Is the Attention Economy Noxious?” Philosophers’

Imprint, 20, no. 17: 1-13.

D’Angelo, Jonathan, and Toma, Catalina. 2017. “There Are Plenty of Fish in the Sea: The

Effects of Choice Overload and Reversibility on Online Daters’ Satisfaction With Selected

Partners”. Media Psychology, 20:1, 1-27.

de Vries, Bouke. 2023. “Selling visibility boosts on dating apps: a problematic practice?” Ethics

and Information Technology 25(2), 30.

de Vries, Bouke. 2024. “State-Run Dating Apps: Are They Morally Desirable?” Philosophy &

Technology 37, 30.

Klincewicz, M., Frank, L., and Jane, E. 2022. “The Ethics of Matching: Hookup Apps and

Online Dating”. In The Routledge Handbook of Sex and Sexuality, Brian Earp, Clare Chambers,

and Lori Watson (eds.). Routledge.

Kugelberg, Elsa. 2025. “Dating apps and the digital sexual sphere”. American Political Science

Review. 1-16.

Heino R.D., Ellison N.B., and Gibbs J.L. 2010. “Relationshopping: Investigating the market

metaphor in online dating”. Journal of Social and Personal Relationships 27(4): 427-447.

Hirschman, Albert. 1970. Exit, Voice, and Loyalty: responses to decline in firms, organizations,

and states. Harvard University Press.

Nader, Karim. 2020. “Dating Through the Filters”. Social Philosophy and Policy. 37(2): 237-

248.

Nader, Karim. 2024. “The gamification of dating online”. Theoria. 1-17.

Pew Research. 2019. “The Virtues and Downsides of Online Dating”

https://www.pewresearch.org/internet/2020/02/06/the-virtues-and-downsides-of-online-dating/

Pew Research. 2023. “From Looking for Love to Swiping the Field: Online Dating in the U.S.”

https://www.pewresearch.org/internet/2023/02/02/from-looking-for-love-to-swiping-the-fieldonline-dating-in-the-u-s/

Pronk T.M., and J.J.A Denissen. 2020. “A Rejection Mind-Set: Choice Overload in Online

Dating”. Social Psychological and Personality Science 11(3): 388-396.

Rosenfeld, M.J., Thomas, R.J., and Hausen, S. 2019. “Disintermediating your friends: how

online dating in the United States displaces other ways of meeting”. Proceedings of the National

Academy of Sciences, 116(36): 17753–17758.

Sandel, Michael. 2012. What Money Can’t Buy: The Moral Limits of Markets. Farrar, Straus, and

Giroux.

Schwartz, Barry. 2016. The Paradox of Choice: Why More is Less. HarperCollins.

CHAPTER TWO: ATTENTION, VIRTUE, AND THE ATTENTION ECONOMY

I. Introduction

Imagine there is a miniature creature living in your pocket. At any time of day, this

creature might interrupt what you are currently doing to alert you to something else it thinks you

would find interesting. It is very clever at doing this because it observes you regularly and has

learned what you tend to find interesting; and it is very hard to ignore—in addition to being loud

and flashy, the things it alerts you about are often very important (like messages from loved

ones) or pleasurable (like a funny anecdote).

This creature has been living in your pocket for so long that you feel like you can’t live

without it. And this is true in a sense—basically everyone else has one and uses them to

communicate with you; they expect you to have one as well; it is socially necessary. You have

also grown habituated to all the ways in which you interact with it and the services, such as

information, that it provides you. It is often the first thing you interact with in the morning and

the last thing you talk to at night. One could add more details to the story, but we can now ask:

would the people in this world, including you, live better lives than in a creature-less world?

Obviously, this story is a thinly veiled allegory for our actual relationship to our

smartphones and the attention economy (AE) products and services they provide.1 A key

question it raises is whether AE contributes to or inhibits our flourishing as human beings. An

1 The attention economy is a large-scale market in which people exchange their attention for access to a new media

service or product (e.g., Facebook). These companies then sell this attention (and the data it generates) to

advertisers.

overall judgment on this question may seem daunting, so I’d like to raise a more specific

question. Namely, does AE interfere with certain capacities or habits that play a central role in

living well? When it comes to those that involve attention, I believe the answer is “yes”.

In section II, I argue that some key forms of attention, which are part of living a

flourishing life, are deeply at odds with AE. I focus on sustained attention, but I also consider

mind wandering and flow. I begin by explaining how there is good reason to believe that AE

interferes with people’s capacity for sustained attention. Then, I argue that sustained attention is

essential for the virtues of aloneness and stillness, as well as a valuable tolerance for loneliness

and boredom. Aloneness and stillness are dispositions to be present in the moment, with oneself

or those close to one, without distraction or interruption. I argue that these are valuable both

intrinsically and because these dispositions help us to realize important values, such as long-term

commitments as well as nuanced self and relationship knowledge. Given that AE inhibits the

capacity for sustained attention, my argument shows the extent to which AE negatively impacts

these virtues and values.

In section III, I examine additional ways that AE negatively impacts human flourishing

by misdirecting people’s attention. I argue that AE encourages “telic” (external goal directed)

activities, particularly competition for attention-based positional goods (e.g., the most “likes” or

“followers”). I argue that this zero-sum positional competition for attention, fostered by social

media, damages individual well-being and social relationships.

II. Sustained Attention and Virtue

A) Interference with Sustained Attention

Here, I argue that there is good reason, both conceptually and empirically, to believe that

AE hinders our ability to exercise and develop the capacity for sustained attention (as well as

mind-wandering and flow). We are constantly on AE apps that are designed to distract us and

divert our attention from one thing to the next. The worry this raises is that by negatively

impacting our ability to sustain attention, AE may inhibit good habit formation—such as one’s

developed ability to be fully present with oneself.

Let’s examine some reasons why it’s likely that AE interferes with the development and

exercise of sustained attention. Consider factors such as the attention-grabbing design principles

of AE services, the profit incentives of AE companies to capture our attention, and the intense

competition all this fosters for our attention. These factors all support the claim that AE is at

odds with sustained attention. It becomes much more difficult to consistently sustain one’s

attention when it is constantly and expertly sought after by many other actors in one’s

environment. Vast amounts of computing power lie behind the AE recommendation algorithms,

and AE sites are designed—via a hooking mechanism that exploits psychological weaknesses—

to keep people using them indefinitely (Castro & Pham, 2020: 2, 7-9).

Castro and Pham explain how the “Hooked Model” of attention capture is designed to

reward users for repeatedly checking their smartphone apps, via triggering and queueing (alerts),

rewarding (providing messages, content, etc. of variable value), and prompting further

investment (e.g., posting in response); this feedback mechanism shares similar key features with

the design of slot machines (2020: 7-8). This provides strong reason to think that the AE

environment has a distracting and diverting effect on one’s attention.

Consider also the practices and social norms surrounding smartphones, which are a key

medium of AE. When an individual is tethered to a device that is persistently sending them

notifications and alerts, both from phone calls and texts, as well as from social media

applications, this can lead to constant interruption from the task at hand. This is reinforced by the

social expectation that one ought to respond relatively quickly to messages from friends, family,

colleagues, etc. Due to the ubiquity of smartphones, one’s response time is expected to be much

shorter now than it previously was for most types of messages, such as phone calls or emails.

Beforehand, one could always say that they weren’t near their phone or computer for a

reasonable amount of time. That excuse is far less plausible given the prevalence of smartphones,

how they are used, and our expectations surrounding them.

Given all of this, there is good reason to think that the adoption of these devices and the

AE-based attention-grabbing apps on them has negatively affected the capacity of individuals to

sustain attention. The habit-forming design of AE services, devised to increase checking—via a

highly interactive device that people are constantly tethered to and that is socially necessary—

likely has a limiting effect on people’s ability to pay sustained attention to anything; this includes

any particular content or activity on the phone itself.

This is further supported by empirical research on the negative impact of AE products

and services on attention spans and sustained attention. A notable source is Ward et. al (2017):

they review some of the literature that supports this causal connection, and they also surprisingly

found that, even when smartphones are not being used, smartphones can reduce one’s cognitive

capacity for other tasks simply by being nearby.

I believe that this provides reasonable support for what I hope was already an intuitively

appealing claim. Going forward I will mainly take AE’s interference with the development and

exercise of sustained attention as a premise, especially in paradigm cases of AE like TikTok.

B) Attention: An Important Component of Virtue

I will now outline why attention plays a key role in living virtuously. A central tenet of

virtue ethics is that to be virtuous one needs to be able to feel and act in the right way. (Aristotle,

NE Bk. II.) I claim that a central part of feeling and acting in the right way is to attend in the

right way to people and things. Importantly, this requires types of attention that are inhibited by

AE—in particular, sustained attention. Sustained attention, which is necessary for deep and

loving attention to oneself and one’s social environment, provides us with the morally and

emotionally salient details needed to respond in a virtuous way. Emotional intelligence requires

proper attention. Consider virtues like generosity and empathy, which require deep and sustained

attention on one’s friends and the intricate knowledge that this kind of attention provides.

Part of this argument rests on the view that sustained attention is analogous to a muscle

that needs to be developed and maintained. This view is rooted in Aristotle’s account in

Nicomachean Ethics, namely, that we develop our natural capacities into virtuous habits and

dispositions by practicing how to use them the right way. Now, some might think that the “right”

way to use our attention is by becoming adept at diverting it and multitasking. However, as the

argument above indicates, and as I will expand upon below, important virtues and values are

diminished or lost by failing to develop the capacity for sustained attention.

Importantly, if I am on the right track, a well-developed capacity for sustaining attention

is essential to developing and maintaining the virtues in general. To further develop this point,

however, I will focus on the specific virtues of aloneness and stillness. In these more concrete

cases, the importance of sustained attention for their development and exercise, as well as the

negative impact of AE, may be easier to see. Moreover, as I will argue, aloneness and stillness

provide important self and relationship knowledge; in turn, this knowledge is needed for

exercising many other virtues (e.g., generosity and empathy). In this way, aloneness and

stillness, which rely on and foster sustained attention, help to underpin the virtues in general.

C) The Value of Aloneness (and Tolerance for Loneliness)

There are specific virtues—aloneness and stillness—that require a well-developed

capacity for sustained attention. Now, one might not think aloneness and stillness are all that

important, but I will explain how they are intrinsically valuable and crucially connected to goods

like self-knowledge and the ability to maintain commitments to people and projects. Moreover, a

well-developed ability to practice aloneness and stillness is deeply intertwined with the capacity

to sustain attention well; they reinforce each other. Because AE interferes with our ability to

sustain our attention, and thus to successfully develop and practice aloneness and stillness, it gets

in the way of values and virtues that are needed to live well.

Consider aloneness: I take it to be the well-developed disposition to be fully present with

oneself, without distraction or interruption. Stillness is similar to aloneness, except that stillness

(as I will use this term) centrally involves being fully present with others. Now, aloneness is

different from mindfulness, which Kabat-Zinn (1994: 4) defines as “paying attention…on

purpose, in the present moment, and nonjudgmentally”. This is because, with aloneness, the

attentional focus is on oneself and one’s experiences, and it can be judgmental. In this way,

aloneness can provide different types of value than mindfulness, such as improving oneself by

actively challenging unhelpful thoughts and feelings as they arise. Moreover, mindfulness is a

state, but I am taking aloneness to be a disposition to be in a certain state (i.e., as a character

trait). However, in the analysis below, I will also use “aloneness” to refer to the state engendered

by this disposition.

It is important to contrast the state of aloneness with that of loneliness. Loneliness is

different, because it involves a negative feeling of disconnection from other people and from the

outside world. Loneliness is about a loss of genuine or satisfying connection to other people. One

can be fully present with oneself (i.e., practicing aloneness) and not feel this type of

disconnection from others. Aloneness is different in this key respect from loneliness.

One reason the distinction between aloneness and loneliness is important is that, from the

perspective of flourishing, loneliness is not a good candidate for being an intrinsically valuable

state or virtuous disposition. Instead, it involves feeling disconnected from one’s social world

and dissatisfied, which (all things equal) would seem to make one’s life go less well.

In addition, loneliness is currently endemic in our society today, and people often refer to

it as a public health crisis. In fact, Surgeon General Dr. Vivek Murthy released an advisory in

2023 stating that loneliness is an “epidemic”.

2 Given this, it would be strange if AE were at odds

with loneliness, because AE pervades our everyday lives. Two things that are at odds with each

other are unlikely to co-exist to such a large degree. In fact, many scholars have argued that AE

products and services have directly contributed to an increase in loneliness. (See e.g., Castro &

Pham, 2020.) Not only do they co-exist, but they seem closely connected to each other.

It is worth noting, though, that the increase in loneliness need not have been accompanied

by increased tolerance for being lonely. There is reason to believe that not only are people

2 https://www.hhs.gov/surgeongeneral/reports-and-publications/connection/index.html

lonelier but that they may be less okay with feeling lonely. Why might this be? Through social

media, AE promises constant social connection, which, when robust connection is not achieved,

may increase feelings of loneliness while (ironically) inhibiting the ability to tolerate them. It is

harder to be okay with feeling lonely when everyone else on social media apps seem so

connected and happy—one is generally shown a curated “highlight reel” of others’ lives.

Because of social media’s AE business model, it is no accident that this highlight reel is also

very attention grabbing.

AE provides the appearance of increased social connection (e.g., via followers and likes)

but many of these are not genuine or satisfying ones, which engenders feelings of disconnection;

and it enhances this sense of loneliness by presenting an image of others as socially wellconnected. The constant and negative social comparison fostered by these AE services could

make feeling lonely less tolerable for many people who use social media.3

Now, it may be that loneliness, while it is not in itself a constituent of living well, can

(say) help to engender creativity in the arts or enhance one’s empathy. This suggests that one’s

relationship to loneliness has instrumental value in the following sense: it may be good to be able

to tolerate loneliness for the valuable things that it helps one achieve, including deeper creativity

and interpersonal understanding. Importantly, the activity of being alone with oneself can

sometimes temporarily slip into loneliness; so, to enjoy and cultivate aloneness, we need to be

able to tolerate loneliness. For these reasons, loneliness per se is not valuable, but the ability to

tolerate it is. But AE increases feelings of loneliness while inhibiting the ability to tolerate it,

which is the exact opposite of what would be conducive to human flourishing.

3 In section III, I will examine how this issue of social comparison connects to positional competition.

Now, having distinguished it from loneliness, let’s return to aloneness. There are many

reasons to think that aloneness—being fully present with oneself—is a valuable state to

experience, and that to live well one should cultivate the capacity to access and maintain it.

Aloneness provides space for important aspects of self-awareness. These include critically

reflecting on one’s experiences, desires, and values, as well as hearing one’s “inner voice” (or

inner guidance). Along these lines, it involves paying attention to anything we are trying to

distract ourselves from, such as things that are painful, hard to engage with, or important to us.

All of these activities require a certain amount of psychological separation from environmental

factors that distract and divert our attention. Moreover, if aloneness becomes a habitual activity

that is consistently done well, it is a plausible candidate for an inherently enjoyable one. Skillful

aloneness exercises central human capacities, including self-awareness and reflection, and this is

intrinsically valuable.

Relatedly, a well-developed capacity for aloneness plays a crucial role in developing selfknowledge. Without skill at being alone, and a distraction-free environment, it would likely be

difficult to critically reflect on and pay close attention to our thoughts and feelings. And without

being able to reflect in this way, a person would not have full access to important information

about their inner lives. Importantly, having adequate self-knowledge is arguably an important

component of living well. For instance, we would struggle to make good decisions if we did not

adequately know who we are and what we really want.

But a capacity for aloneness involves the ability to sustain attention, which AE hampers

due to its structure, purpose, etc. The moral concern is that AE regularly distracts and diverts our

attention, which turns us into people that can’t focus on ourselves in depth or detail; and hence

AE negatively impacts people’s ability to cultivate and maintain the virtue of aloneness.

To this concern, it may be objected both that people’s desire to distract themselves is

nothing new and that distraction is not always a bad thing. In fact, it seems like losing oneself in

(say) a hobby or work may help to keep one from excessive rumination on one’s troubles.

My first response is as follows. I’m happy to grant that this urge and tendency to

distraction from oneself predates AE—people have been fleeing from “knowing themselves” for

a very long time. It seems plausible that people have always had trouble handling the boredom,

anxiety, and discomfort that often accompanies inner reflection. In fact, we see recognition of

this human psychological tendency in the works of poets, philosophers, etc. from earlier periods.

For example, Nietzsche criticized the habits and values of his fellow 19th century

citizens—e.g., an obsession with newspapers, watches etc.—in works like The Gay Science,

where he laments that “One thinks with a watch in one’s hand…one lives as if one always ‘might

miss out on something’”. (1974: 259) In this section (329), he suggests that, by needing to be

perpetually busy, his peers had failed to develop an ability and appreciation for “prolonged

reflection” and that there had been a corresponding diminishment of the value of leisure and

contemplation. (1974: 259-260) The desire to avoid sustaining attention on oneself is not new.

However, even granting that the tendency toward distraction is not new, I argue that the

AE turns a temptation to distraction into a constant and pervasive habit and environmental

ecosystem. AE seems precisely designed to distract us from ourselves. It is like a supercharged

version of Nietzsche’s critique of the attentional environment of his era. The amount of

distraction from oneself that AE makes available, and the power and reach of the technology

supporting it, is substantially greater than it was in Nietzsche’s late 19th century. AE reinforces

our tendency towards distraction to a much larger degree, and because of it, we now live in a

much more distraction-prone world.

Moreover, even if some distraction is good, too much distraction gets in the way of other

important activities, such as adequately being with oneself and reflecting on one’s inner

experience. Relatedly, distraction might be helpful in the short run to feel better, e.g., by

alleviating anxiety, but it can backfire if it becomes one’s primary long-term strategy and habit.

In this case, one wouldn’t get enough practice learning how to tolerate loneliness and cultivate

aloneness. But AE relies upon and promotes precisely this way of approaching our lives—

namely, dealing with it by constantly distracting ourselves. In an AE environment, the ratio of

distraction to sustained attention is not well balanced.

Now, perhaps AE’s distraction ecosystem is ultimately a difference in degree rather than

in kind from what things were like beforehand. But this would still amount to a very substantial

difference. Importantly, it inhibits aloneness to a greater extent, which places this virtue much

further out of reach and makes knowing oneself even more difficult.

My second response is that there are better and worse ways to distract oneself. For

instance, there is a qualitative difference between hobbies that engage many key capacities (e.g.,

playing guitar) and mindless scrolling through one’s social media feed. When one gets lost in the

former type of activity, one’s attention is diverted away from oneself and one’s troubles, but

one’s attention isn’t fragmented and diluted in the way it is during repetitive social media

scrolling. Moreover, the mental state one experiences while playing guitar can be much closer to

one of flow (or perhaps mind-wandering), whereas social media scrolling is too disjointed and

externally directed an activity to lend itself well to these pleasurable and valuable mental states.

(I will have more to say later on about flow and mind wandering.) When the activity consists of

constant interruption (as it does with social media), it won’t be as relaxing and rejuvenating as

uninterrupted downtime with a favorite hobby (e.g., playing guitar).

D) The Value of Tolerance for Boredom and the Value of Stillness

Having considered aloneness, we can now examine the riveting topic of boredom, while

keeping track of stillness, a virtue that relates to tolerance for it. What is boredom? For our

purposes, I present a mostly stipulative account. I hope that it is plausible, but ultimately what

matters is that the phenomenon I’m describing plays a role in living well. First of all, boredom is

a type of dissatisfaction. Moreover, a central component of boredom is not having one’s desire

for excitement or stimulation satisfied. Presumably, people often have such a desire to feel

excited, entertained, or gripped by their experience and activities. When one is bored, one does

not feel excited or engaged by anything in that moment, and one is dissatisfied about this fact.

Now, consider how AE seems designed and structured to enable people to avoid

boredom. If the goal of a site or company is to capture people’s attention, then they will want

their users to be as stimulated as possible. If people using a site become bored, the company runs

the risk of losing their attention (both temporarily and permanently)—and hence their source of

revenue. This is because excitement and interest are what drive further user engagement (e.g.,

clicking on links or scrolling through feeds). This is a clear and powerful motive to avoid having

bored users at all costs, as well as to lower their tolerance for boredom.

However, I argue that tolerance for boredom is valuable for the following reason.

Tolerance for boredom—being comfortable with lack of excitement or entertainment, even when

one wants to feel more excited—is needed for realizing important things like commitment and

self-knowledge, as well as stillness. In this way, developing a tolerance for boredom is a virtue.

Boredom is connected to commitment, which centrally involves being able to see things

through. Because seeing things through, especially long-term projects, involves periods of time

when one is not gripped or excited by what one is doing, commitment is a value that can’t be

fully realized without some amount of boredom and toleration for the discomfort this brings. To

take an example familiar to academics, imagine you have a long-term writing project. Guiding

oneself successfully through this type of writing project will almost certainly involve many

periods of dissatisfaction, including the absence of feeling gripped (i.e., boredom).

One idea underlying this explanation is that commitment can be deeply at odds with

short-term preference satisfaction (e.g., for excitement). In this case, commitment to a long-term

project is at odds with the short-term preference to be distracted (which, in turn, stems from

discomfort with boredom and a desire to avoid it).

Let’s look at another important type of commitment, namely the one involved in

maintaining close, intimate relationships with friends or family. In this case, boredom, including

accepting quiet moments and lack of excitement, may be a key part of intimacy. In this way,

intimacy (via commitment) seems to require tolerance for boredom, including being able to stay

focused on one’s partner when some (or all) desires for excitement are going unsatisfied.

Otherwise, we run the risk of being inconstant towards the object of our commitment. No person

or project will always satisfy all of our desires about them. Boredom is a type of dissatisfaction

and, given that projects and relationships sometimes engender boredom, tolerance for it is

necessary to seeing them through.

My argument shares similarities to Pettit’s (2004: 161) account of the value of hope; he

argues that hope serves to bracket negative thoughts by keeping one’s focus, especially during

challenging times, on the possibility that things will turn out okay. What I am arguing is that

commitment requires bracketing (temporary) negative feelings about a person (or project) to

maintain one’s commitment. Tolerance for boredom is needed to bracket these feelings.

Further, when it comes to the ability to form and maintain commitments, especially

within close relationships, attending to someone deeply is arguably a key part of what loving

them involves (cf. Murdoch, 2001). An example of potentially improper attention is the

widespread phenomenon of people checking their phones at dinner tables, at social gatherings,

etc. Being habituated to desire and expect constant excitement—and hence having a sizeable

discomfort with boredom—is driving a fair amount of this behavior.

The general point is that we’re always connected to AE, but that commitment in intimate

relationships (or long-term projects) requires paying focused and sustained attention to who you

are with or what you are doing—even, perhaps especially, when you are bored or dissatisfied.

We must be able to weather such intermittent periods of boredom. (Note that this point is

compatible with persistent boredom being a reason to give up a relationship or project.)

In this way, commitment requires a separate local space from the outside world and the

interruptions it engenders, but this is hard to get when the entire outside world is “in your

pocket”. Without tolerance for boredom, commitments become much harder to realize or

maintain; but AE is set up to capture your attention by eliminating boredom, and so we don’t

develop this tolerance. This is the crux of the problem. Smartphones and their AE model take

away our capacity to weather boredom.

Relatedly, stillness, which is like boredom but without the dissatisfaction (regarding

one’s state of non-excitement), requires the capacity to sustain attention on social experiences

that are not constantly stimulating. Deep and sustained attention on another person will likely

sometimes be boring and at it certainly won’t always be exciting. Someone with a disposition for

stillness would be skilled at managing such moments of non-excitement.

Importantly, stillness is something one can start to enjoy and feel at home experiencing.

This is like enjoying a quiet moment with one’s friend or lover without the need to do or say

anything exciting. These types of calm moments shared with a loved one are enjoyable and

valuable in themselves, and they should be cultivated and enjoyed. Moreover, since some nonexciting moments are bound to occur in any relationship, cultivating the ability to experience and

enjoy stillness in relationships would be an aid to stronger commitments.

Finally, being fully present with oneself and adequately attending to one’s internal

experience involves a certain amount of tolerance for discomfort with a lack of excitement or

stimulation. That is, skillful aloneness also requires tolerance for boredom. Importantly, this

explains why tolerance for boredom is required for self-knowledge—namely, this tolerance is

needed for aloneness and, hence, aptly attending to oneself. Aloneness and tolerance for

boredom are thus intricately tied together.

I endeavored above to highlight how aloneness and stillness are intrinsically valuable—

they are states that would be good for us to learn to appreciate—and that cultivating these virtues

requires a tolerance for loneliness and boredom.

I also argued that aloneness and stillness, along with tolerance for loneliness and

boredom, play a key role in commitment and self-knowledge. I have been mainly presuming that

commitment and self-knowledge are valuable components of a life well lived. Although I think

this is quite plausible—that significant value is realized by (say) engaging in and completing

long-term projects—perhaps some may disagree. However, my aim here was primarily to

elucidate an important connection between aloneness and stillness, a tolerance for loneliness and

boredom, and these further values; and to show how, by inhibiting sustained attention, AE

wrecks all of them. This part of my argument demonstrates that to the extent that commitments

and self-knowledge are valuable, we have additional reason to be wary of AE.

E) Other Attentional Virtues: Mind-Wandering and Flow

There is also substantial value in the inverse of sustained or focused attention, which I

will call “mind-wandering”. This involves the absence of the directed aspect of attention.

Allowing one’s mind to wander is like letting your attentional muscles rest. Mind-wandering is

connected to insight. For example, sometimes you focus hard on a problem, then take a break by

(say) aimlessly taking a walk, and then additional ideas come to you. Rest and inspiration are

some valuable aspects of mind-wandering.

Another important attentional state is “flow”—both research and common sense suggest

that flow is valuable both for the deep enjoyment of the activity it provides and the enhanced

productivity that results from it.4 Like mind-wandering, there is a restorative and insight eliciting

nature to certain flow-based activities—e.g., when one is fully absorbed in fishing, driving, etc.

And like mind-wandering, its phenomenology seems to involve “turning off”, or disconnecting

from, the self-conscious part of one’s attention.

Yet, mind-wandering is different from flow in the following respect—flow can still occur

when one is actively directing one’s attention (at least to some extent). Take for instance a

musician or athlete who is in a flow state. It would be strange to suggest that they lack focus on

the task at hand. If so, they are not in the state of mind-wandering. This distinction between

mind-wandering and flow is present in the psychology literature. In fact, Deng et. al (2022)

found that these two states are inversely correlated.

4 The original work on the psychology of flow was done by Mihály Csíkszentmihályi. See his book Flow: The

Psychology of Optimal Experience (1990).

Let’s return to the connection between AE and attentional virtues. What is crucial for our

analysis is that AE is at odds with both mind-wandering and flow. These forms of attention,

which are valuable for the above reasons, cannot be fully cultivated when one is constantly

engaging with AE products and services. Again, this is due to the addictive, attention capturing

nature of AE.

Instead, the form of attention AE encourages is a sort of in-between attentional state—we

are neither in a state of mind-wandering nor absorbed in flow. (As a guiding example, recall the

difference between (say) playing guitar and scrolling through a social media feed.) This is

because interacting with AE services like a social media feed drives us to flit from one thought

(one object of attention) to the next, in an unintentional and disordered way. This attentional

process is clearly in tension with the uninterrupted attention required by flow (e.g., the flow state

one can attain while playing guitar).

Importantly, while this attentional state is disjointed and not directed from within oneself,

it is not free from any control. So, it is at odds with letting one's mind wander. There is too much

external noise, change, manipulation, etc. for the mind-wandering process to occur. Unlike mindwandering, the thoughts, ideas, etc. that pop up or are attended to are not one's own. They are not

the result of any kind of internal process of “mental digestion” of items previously focused upon,

as they are in the subconscious and internal process of mind-wandering. In sum, there is a

tension between AE and the valuable states of flow and mind-wandering; this is in addition to its

tension with sustained attention.

III. AE and Positional Goods

I argued above that AE impedes human flourishing by degrading people’s attentional

capacities and hence their ability to live virtuously. In this section, I provide an additional reason

why AE is in tension with human flourishing: namely, that AE encourages telic activities,

particularly competition for attention-based positional goods, which in turn degrades individual

well-being and social relationships.

Now, one common criticism of AE products and services is that they cause mental health

problems, such as anxiety and depression, especially in younger people. Jonathan Haidt (2024)

notably defends this claim in his book The Anxious Generation; and there is other empirical

support for a causal connection. For instance, Fruehwirth et. al (2024) found that adding an hour

of social media use had a significantly negative effect on “the mental health of college students

during the Covid-19 pandemic”. Moreover, there is enough evidence of social media’s negative

impact that the Surgeon General advised in 2023 that it may not be safe for young people’s

mental health. 5 6 If there is a causal connection, then AE is harmful to individuals due to its

negative impact on this aspect of well-being. (cf. C&P, 2020: 3)

I hope that my moral analysis below helps illuminate what may underlie this connection.

However, my argument that AE misdirects our attention in a way that damages individual and

5 https://www.hhs.gov/surgeongeneral/reports-and-publications/youth-mental-health/social-media/index.html

6 Additionally, drawing on a wide body of research, especially Twenge (2017), Castro and Pham argue that

“Consumption of new media causes anxiety, depression, feelings of loneliness…and the attention economy is

implicated in these concerns.” (C&P, 2020: 3, my emphasis).

First, they note that many studies show that new media consumption and diminished mental health

outcomes are correlated. (C&P: 3) They then cite studies showing that decreasing Facebook use increases overall

psychological well-being; and they also point out that new media use keeps people from activities, such as exercise,

that have been shown to increase well-being. (C&P: 4) Both of these findings suggest a causal connection between

new media use and subsequent negative mental health outcomes. Finally, C&P observe that a dramatic “spike in

mental health problems” coincided with the ubiquitous adoption of smartphones. (C&P: 4) C&P argue that these

points strongly suggest a causal connection from the use of AE services to increased mental health problems.

social flourishing would still be a concern even if AE did not have a direct and negative causal

impact on mental health.

A) Atelic vs. Telic Activities

A distinction that plays a key role below is between telic and atelic activities (drawing on

Setiya’s “The Midlife Crisis”). According to Setiya (2014: 12-13), atelic activities are those that

“do not aim at a point of termination… [such as] going for a walk” or living well, while telic

activities are those that aim at “a terminal point… [such as] buying a house [or] getting a job.”

These activities end once the house or job is obtained.

This distinction involves the reason or motivation behind engaging in an activity. Telic

activities are those that are done for a reason or goal that is beyond the momentary action itself

and that results when the activity is finished. Think of paying one’s monthly bills—this is done

for the purpose of ensuring one’s basic needs are met. It is goal directed; it is a task that can be

completed (and also repeated ad nauseum). On the other hand, atelic activities are those which

are done for their own sake. An atelic activity is one in which the person engaged in it has no

further reason for doing it. For example, a person may sometimes play their guitar for no other

reason than to do the activity itself. It is already complete in the moment it is performed (in that

there is no further goal to achieve). In cases like this, the person is “not on the way to achieving

[their] end. [They] are already there” (Setiya, 2014: 13).

Of course, people often have mixed motivations for performing an activity. For instance,

a person may play their guitar both to do it for its own sake (intrinsically) and to try to gain

social approval. So, it is sometimes hard to determine, even from a first-person perspective,

whether an activity is telic or atelic. For this reason, it might be more accurate to call activities

atelic when they are primarily engaged in for no further purpose; where the driving motivation is

to do the activity itself, not to complete it by reaching some endpoint or goal.

Although Setiya focuses on Aristotle’s Metaphysics, the atelic vs. telic distinction is also

connected to Aristotle’s discussion of eudaimonia in Books I and X of Nicomachean Ethics.

Aristotle argues that happiness must be an activity that is done for its own sake (and nothing

else), and that is complete in itself. Specifically, happiness is rational activity—both practical

and theoretical—performed excellently or virtuously, over the course of a lifetime (NE, Bk. I.7).

So, for Aristotle, the best type of activity (the life of reason) is one that is done for its own sake

and that has no end point. This means that a flourishing person is, to the extent possible, engaged

successfully in specific atelic activities. However, atelic activities are a broader category because

they need not be the specific rational activities that constitute happiness for Aristotle.

With this distinction in mind, let’s examine why AE, including the social media services

that utilize this economic model, may be degrading people’s ability to live well.

B) Telic Activity: Competition over Positional Goods

One explanation is that people are using social media and AE in a telic fashion. Now,

there are different types of telic activities that people may be engaged in on social media.

However, a key one for this analysis is that people may be engaged with these AE-based

products in order to gain or maintain popularity and social status. This is a telic activity because

social media is used for the goal of improving one’s social standing—there is a terminal point at

which the person is aiming. It is also often instrumental, in that social status is not valued in itself

but for the rewards it brings.

Importantly, this is a telic activity because it aims at a certain end, namely, increasing

one’s popularity by (say) accumulating more and more likes or views—what is focused on is the

goal (e.g., gaining more likes) not the process or activity itself. In this way, it is a special type of

telic activity, namely, one where a short-term goal is aimed at and focused upon (e.g., increasing

one’s likes today), even when the ultimate terminal point may be unreachable. (cf. Setiya, 2014:

12.) This is because (practically speaking) one can always be more popular. While popularity is

the final aim of the activity, a person works towards it via a chain of short-term ends.

To the extent that people are engaged with AE in this way, there is a clear case for why

this would have significant moral downsides. Telic activities already have downsides: anxiety (or

at least dissatisfaction) before the goal is achieved—because you don’t yet have what you

want—and emptiness (and more dissatisfaction) after it is—because there is no longer something

to strive for. (cf. Setiya, 2014: 11-12)

In addition to this, vying for popularity is a telic activity whose end is particularly hard to

achieve and sustain—the number of “winners” (popular people) is inherently limited. For many

people, even the (momentary) satisfaction of completing this activity by achieving more

popularity is not obtainable. As we will see, popularity (or social status) is one type of what is

called a positional good.

What is a positional good? It is roughly characterized as follows. The value of a

positional good is its relative value. In particular, it is a good whose value depends on its relative

ranking—its rank in an ordering—with other goods of the same kind. (See Brighouse and Swift,

p. 472) This is rather abstract, so it is helpful to look at a few examples. Consider car size. (See

Halliday and Thrasher, p. 175-176) While the size of one’s car delivers multiple goods, one way

it provides value is by enhancing safety. But the amount of safety provided by size is primarily

due to its relative size. That is, a car’s safety value depends on the size of other vehicles. To

determine whether a car’s size will keep one safe, we need to know some additional context—

namely, how big the other cars are. Compact cars are much less safe if everyone else is driving

an SUV (and much safer if other people are driving even smaller cars). For car safety purposes, it

is relative size that matters.

Or consider number of votes. The effective value of having one vote depends on the

number of votes that other people have. If everyone else is given a second vote (when before

they had just one), then my one vote automatically becomes less valuable. Contrast this with the

number of sandwiches. The value of having one sandwich doesn’t typically depend on how many

sandwiches other people have. With votes, the relative number (i.e., positional rank) is crucial;

with sandwiches, the absolute amount is what matters.

It is important to point out there are many reasons why people engage in positional

competition; it is often not about envy or jealousy (cf. Frank, 2011), but about access to

important goods. Positional goods may be desired not to signal a higher status or “do better” than

others; but instead to maintain some pre-existing good, whose value depends on what others are

doing. (See Lichtenburg, 1996.) For example, one might be fine with the number of bathrooms in

one’s house; but if everyone else adds a bathroom, one is pressured to do so as well (to maintain

the value of one’s home).

Let’s return to AE and social media. Certain social media platforms rely upon and

encourage thinking of social relationships in a telic and positional fashion. Social media’s

features (e.g., likes) invite competition for attention, status, and “influence”; when engaged with

in this way, this may undermine one’s flourishing because it excessively redirects one’s focus

towards one’s positional status. Social media is ubiquitous, and many people have a curated

social media page (e.g., Instagram). These sites encourage people to constantly think about social

status—their attention-based algorithms work in tandem with our basic desire for social

approval. Almost everyone is in an arms race to appear the happiest on social media. But there

can only be a limited number of winners, since social popularity is a positional good.

In the case of AE, some people are using social media to compete over the positional

good of popularity or social status. This may in turn encourage others to compete, e.g., to not

lose ground. Because popularity is a positional good, many people will “lose out”—they will not

gain the status or popularity they were trying to achieve. Think of musical chairs with many

people and only a few spots—only a small subset of people can achieve popularity (get the

limited spots). (cf., Frank, chapter 2.) For those who don’t achieve their telic goal, this will be

frustrating. And the competition itself likely produces stress and anxiety. For one thing, to the

extent that it is valued instrumentally, people won’t enjoy the activity itself (since they would

only care about the positional result). Also, it is an ongoing positional competition, where, as I

will expand upon below, unlike other competitions (e.g., card players in a pub) one is induced to

instrumentalize one’s social relationships to achieve their ends; and AE is set up to constantly

alert people to threats or opportunities relating to their social status.

When people engage with social media in this way, it leads to problems like frustration.

Recall that Castro and Pham (2020: 3-4) discuss many studies which show that social media use

likely causes mental health problems (e.g., anxiety) in teenagers. Plausibly, social comparison is

playing a key role here. People are vulnerable to social status seeking, and social media redirects

one’s focus toward tangible social comparisons (e.g., likes or “followers”) and away from any

non-competitive aspects of our social lives.

There are three main problems here: dissatisfaction before one reaches their goal (e.g., a

relatively high number of likes); a lack of clarity of what to do once one reaches it; and few

people can even achieve the goal because popularity (as measured by (say) having a relatively

high number of likes) is a positional good. Hence, the telic activity of positional competition for

popularity on social media will tend to be particularly unfulfilling and at odds with living well.

Now, the case of AE and social media fits into a larger explanatory framework for why

positional goods are morally problematic. Explicating this framework will help further elucidate

what’s wrong in this particular case. For starters, positional goods tend to be socially-based,

competition for them often engenders arms races (e.g., military build-ups), and it requires extra

effort just to stay even—"rat-races” in (say) the amenities in one’s house (Hirsch, 1976: 53, 67).

Specifically, the zero-sum games comprising competition for positional goods (e.g.,

social status or popularity) are morally problematic. The more of a positional good one person

has, the less someone else does (Hirsch, 1976: 52); and often they are the object of ongoing

competitions. These are the key problems that underlie the arms and rat-races. They tend to

occur as long-term processes. (cf., Halliday and Thrasher, 2020.) This leads to a potential waste

of resources, including money and energy, because the group (and most or all individuals

involved) is not better off as a result of this competition—it does not increase the supply of the

desired good, because supply of positional goods is inherently limited. So, competition for

positional goods is thus often a large resource sink and one that sometimes diminishes other

connected goods7

. This is one main way in which positional goods are morally problematic.

7 Positional competition sometimes decreases the amount of a connected absolute good such as safety; this is often

the case in head-to-head evolutionary competitions. Positional competition for larger antlers among bull elk, which

temporarily gave individual elk a relative advantage in mating, also led (as the evolutionary competition went on) to

bull elk being less able—as individuals and a group—to avoid predators due to their heavier antlers; this result was a

disadvantage in terms of absolute safety. (Frank, 2011)

In the case of popularity (or social status), as a matter of positional logic, only a certain

subset of a social group will be able to obtain that good. No matter how much competition

occurs, with respect to the good of popularity, the group will be no better off; and most (or all)

individuals will have invested a lot of energy in “getting nowhere” (relative to other people).

The frustration, futility, and stress of this telic activity, with its generally unachievable

goal, is at odds with flourishing; and it may also help to explain the connection between AE and

depression and anxiety. For one thing, there are fewer upsides to counterbalance the negative

aspects of this activity. That is, competition for overall popularity on social media has fewer

obvious redeeming qualities than (say) competitive sports, which involve increased physical

health, potential excellence in a particular skill and practice, and a robust sense of comradery.

In addition, the self-interested competition positional goods require risks tearing at

communal bonds and inculcating bad attitudes about social relationships—in that they become

narrowly self-interested and overly competitive. This is because, in zero-sum games, we can’t

both win—there is no cooperative solution or collective good to be realized. To return to our

case of social media, by promoting a focus on social status (via likes, etc.), AE may shift

attitudes about the purpose of social relationships in a more instrumental direction, and it may

tear at social bonds by making social life more competitive, positional, and quantified.

When people use social media in a telic, positional fashion, as most popular platforms

encourage (e.g., by making certain information salient, such as numbers of likes and followers),

this leads people to view their social lives from the positional standpoint. This positional

standpoint emphasizes what’s comparable, ignoring any values irrelevant to positional

competition. It encourages an unhealthy degree of the economic way of thinking, including

narrow self-interest (friends are for boosting status) and transactional attitudes towards social

relationships (e.g., people trade likes). Competition for positional aspects of goods focuses

attention on their tangible and often instrumental features (e.g., friendship in terms of Instagram

likes). This corrupts or degrades other important aspects of these social goods, including robust

community and friendships, which would be a significant loss. Let me defend this point further.

From a virtue ethics perspective, a morally problematic feature of AE-based positional

competition is how it encourages people to instrumentalize aspects of their personal lives and

friendships for the external goal of gaining or maintaining social status. The worry here is that

there are better purposes for (say) friendship than instrumental ones (such as likes), but that AE

engendered positional competition pushes people to value friendship in this way. For example,

for a telic AE user (aiming at boosting social status), a day at the beach with friends might be

spent trying to capture just the right picture to maximize their number of likes on Instagram. But

this person will then be less engaged with other aspects of their experience. They will be less

focused on other important atelic aspects of it, such as enjoying their friends’ company and

appreciating the natural beauty, without regard for what will result from these activities later on.

This is just a small example—how a person treats one day at the beach. Is this a problem?

Maybe it’s okay if it only occurs occasionally. But if it becomes a way of life—if most social

situations primarily become opportunities to increase one’s social status—then we have a bigger

problem. In this case, the positional approach to social life starts to crowd out other aspects of it.

Importantly, for those caught up in positional competition on social media, the incentives

encourage exactly this type of attitude and behavior. For instance, almost every social activity

can be thought of as an opportunity to boost one’s social media presence.

Two things are worth mentioning here. Of course, people have engaged with social life in

these types of telic and positional ways in the past. But the problem is that AE relies upon,

promotes, and facilitates such behavior. Second, there is a normative assumption involved—

namely, that there are better and worse ways to engage in certain activities, relationships, etc.,

such as a day at the beach with friends. But it is hard to think of (say) friendship as a morally

neutral category—we want our friends to value us for who we are, not whether we can boost

their popularity. And we think that we have good reason for this; friendship in the latter sense

seems to be a degraded version of it. This idea goes back to Aristotle: in Nicomachean Ethics

Book IX he distinguishes friends of virtue from those of pleasure and use. A friend of virtue, in

which there is mutual appreciation of each other’s good qualities, is a higher version of

friendship than a friend of use (one we transact with to gain some external advantage).

To take some other examples of degradation of our attitudes and social life, consider

powerful emotions and important life events. For the former, instead of (say) experiencing grief

or joy without further aims or regard to its external uses, people may start to use these emotions

and the connections they can bring to instead boost their social media status. Relatedly, a

wedding may become just another way to promote one’s own social media presence, instead of

something to be cherished in itself: that is, as a time to be present with the ones you love.

Let’s return to the question of mental health. To the extent that people are missing out on

goods conducive to flourishing, such as robust community and friendship, this can help explain

why AE and social media is implicated in mental health issues. Mental health issues such as

anxiety and depression would be the symptoms of a deeper issue—the diminishment of

important goods, including social ones, that are part of living well. My suggestion here is that the

mental health problems are downstream from the loss of these important goods. However, even

if there is no direct connection to mental health, my account helps to shed light on what is

morally problematic about AE. Namely, by encouraging and relying upon positional

competition, it is inherently in tension with important aspects of living well.

Setiya (2014) argues that we need atelic activities to live well—otherwise we are left with

a sense of meaningless in our lives, with the feeling that we are always chasing further ends. I’ve

endeavored to show that, not only is positional competition on AE a damaging telic activity, but

that it also takes our attention away from meaningful atelic aspects of our lives, including simply

being with our friends and family.

IV. Conclusion

In section II, I argued that AE is deeply at odds with certain forms of attention that are

essential to living a flourishing life. I focused on sustained attention, but I also examined mindwandering and flow. I began by highlighting how AE interferes with sustained attention, while

outlining how a well-developed capacity for sustained attention is needed to be fully virtuous.

Then, I argued that AE inhibits developing a capacity for aloneness and stillness, along with a

corresponding tolerance for loneliness and boredom; and that these are virtuous dispositions both

intrinsically and because they—as well as a tolerance for loneliness and boredom—are required

for realizing important values such as self-knowledge and commitment.

In section III, I developed an additional explanation for the negative impact of AE on

human flourishing. I argued that the attention-based positional competition for popularity and

social status, fostered by social media and its AE model, damages individual well-being and

social relationships. I argued that this zero-sum competition is a particularly unsatisfying form of

telic activity; and that it relies upon and promotes an instrumental view of social relationships,

such as friendship, while taking away from meaningful atelic aspects of people’s social lives.

REFERENCES

Anderson, Elizabeth. 1993. Value in Ethics and Economics. Harvard University Press.

Aristotle. Nicomachean Ethics in Readings in Ancient Greek Philosophy 5th Edition. ed. Cohen,

Curd, and Reeve, C.D.C. Hackett, 2016.

Brighouse, Harry, and Adam Swift. 2006. “Equality, Priority, and Positional Goods”. Ethics,

Vol. 116, No. 3: 471-497.

Castro, Clinton, and Adam Pham. 2020. “Is the Attention Economy Noxious?” Philosophers’

Imprint, 20, no. 17: 1-13.

Csíkszentmihályi, Mihály. 1990. Flow: The Psychology of Optimal Experience. Harper & Row.

Deng Y-Q., Zhang B., Zheng X., Liu Y., Wang X. and Zhou C. 2022. “The impacts of mindwandering on flow: Examining the critical role of physical activity and mindfulness.” Frontiers

in Psychology 13.

Frank, Robert. 2011 The Darwin Economy: Liberty, Competition, and the Common Good.

Princeton University Press.

Fruehwirth, Jane, Alex Weng, and Krista Perreira. 2024. “The effect of social media use on

mental health of college students during the pandemic”. Health Economics, 33(10): 2229–2252.

Haidt, Jonathan. 2024. The Anxious Generation: How the Great Rewiring of Childhood is

Causing an Epidemic of Mental Illness. Penguin Press.

Halliday, Daniel, and John Thrasher. 2020. “Keeping Up with the Joneses (and the Kardashians):

Positional Goods and Wars of All against All”. The Ethics of Capitalism, chapter 9. Oxford

University Press.

Hirsch, Fred. 1976. Social Limits to Growth. Harvard University Press.

Kabat-Zinn, Jon. 1994. Wherever You Go, There You Are: Mindfulness Meditation in Everyday

Life. Hyperion.

Lichtenberg, Judith. 1996. “Consuming Because Others Consume”. Social Theory and Practice

22 (3): 273–297.

Murdoch, Iris. 2001. The Sovereignty of Good. Routledge.

Nietzsche, Friedrich. The Gay Science. Translated by Walter Kaufmann. Vintage Books, 1974.

Pettit, Philip. 2004. “Hope and Its Place in Mind”. The ANNALS of the American Academy of

Political and Social Science. 592 (1): 152-165.

Sandel, Michael. 2012. What Money Can’t Buy: The Moral Limits of Markets. Farrar, Straus, and

Giroux.

Setiya, Kieran. 2014. “The Midlife Crisis”. Philosophers' Imprint 14: 1-18.

Twenge, Jean. 2017. iGen: Why Today’s Super-Connected Kids Are Growing Up Less

Rebellious, More Tolerant, Less Happy—and Completely Unprepared for Adulthood. Simon &

Schuster.

Ward, A. F., Duke, K., Gneezy, A., and Bos, M. W. 2017. “Brain Drain: The Mere Presence of

One’s Own Smartphone Reduces Available Cognitive Capacity”. Journal of the Association for

Consumer Research, 2: 140–54.

CHAPTER THREE: AUTONOMY AND THE ATTENTION ECONOMY

I. Introduction

The attention economy (AE) is characterized by 1) an exchange of our attention for

access to new media services, 2) who then sell our attention, and the data it generates, to

advertisers. (See Castro & Pham (C&P), 2020: 2.) In AE, each person’s attention is used as both

a currency for exchange and as a company’s product and resource. Smartphones and the forprofit apps that run on them (e.g., Instagram) are a ubiquitous part of everyday life for most

people. In 2024, 91% of Americans owned a smartphone phone and 98 % owned a cell phone;

on average, they spent “4 hours and 30 minutes [every day] on their phones… [checked them]

…144 times a day”; and 75 % said that they feel “uncomfortable without their phones”.

Now, many ordinary markets and market exchanges do not violate one’s autonomy.

Consider a person who wants a basic grocery item and purchases it from their local store. In fact,

one common defense of markets is that they respect the autonomous choices of adults (cf.

Nozick, 1974). However, when it comes to AE and the apps that implement it, we might worry

that people “have become tools of their tools”, as Thoreau lamented about his fellow 19th century

New Englanders (1854: 132). AE pervades people’s lives, and because AE services are well

designed to capture our attention, they divert and distract people regularly. One might (say) open

their phone with the intention to quickly connect with a friend, only to end up spending an

unplanned half hour scrolling through their social media apps.

1 https://www.consumeraffairs.com/cell_phones/cell-phone-statistics.html

This raises the question: does AE violate people’s autonomy? If so, why is this the case?

I believe that the answer is “yes”, and in section II, I present several connected reasons for how

and why this is so. First, while authors such as Bhargava and Velasquez (2021) have criticized

AE products and services for being addictive, I argue that—because it inhibits sustained and

directed attention—AE involves a deeper autonomy violation than addiction per se. Then, I

expand my case by arguing that AE is significantly more pervasive, practically necessary, and

hidden than the advertising markets that predated it; that AE interferes with satisfying important

second-order preferences; and that it objectionably inhibits internal preference formation.

Now, if AE violates people’s autonomy, this seems to entail that each person has an

obligation to minimize their use of AE. (cf., Aylsworth and Castro, 2021.) However, people also

seem to have competing obligations to remain on AE; these include social ones like adequately

attending to their friends and democratic ones like staying politically informed. In section III, I

examine these competing attentional obligations, and I explain how AE engenders a tension

between them and the obligation to protect one’s autonomy. I argue that this provides normative

justification for supporting non-AE avenues for socializing and obtaining information.

II. Why AE Violates our Autonomy

A) A Deeper Autonomy Violation than Addiction

Some recent articles critiquing AE have emphasized a connection between addiction and

diminished autonomy. For instance, Castro and Pham (2020: 7-9) argue that a main reason AE is

noxious is that its services are designed to be addictive, and hence that AE’s business model

depends on weakening the agency of its users. Moreover, Bhargava and Velasquez (2021) argue

that social media addiction is common, that AE and its incentive structure play a key role in this

addiction, and that these two features are morally objectionable.

Yet, focusing on addiction generates puzzles about what addiction is, whether it is always

bad, and how exactly it connects to autonomy violations. Moreover, the weight of this critique

depends on whether and to what extent people are addicted to AE services.

I will pursue a different track. I think that there is a deeper sense in which AE violates

people’s autonomy—it interferes with people’s ability and capacity to direct and sustain their

own attention. This, I argue, is a more fundamental violation of autonomy than addiction on its

own, particularly regarding common ones such as addictions to gambling or cigarettes. Here I

move beyond Aylsworth and Castro’s (2021: 664) claim that “smartphone addiction is unlike

other addictions” due to how it “capture[s] our attention”. Our relationship with our smartphones

is not simply a different type of addiction. Instead, I argue that frequent AE use—regardless of

addiction—is a deeper autonomy violation than addiction, and, moreover, that when addiction

does occur, attention capture is partially what underlies it.

AE interferes with one’s ability to control and direct one’s attention. But this capacity is

plausibly a central way that we exercise and express our autonomy. To the extent that control

over our attention and how we direct it is a core part of who we are as agents, this interference

constitutes a serious autonomy violation. For one thing, without this type of internal control,

one’s willpower would be significantly diminished. The ability to sustain and direct one’s

attention plays a key role in monitoring and following through on one’s intentions and plans.

Moreover, addiction to AE media, which involves a lack of attentional control, is likely a

downstream consequence of how AE interferes with this fundamental attentional capacity.

Now, AE products and services are indeed designed to be addictive; they are designed to

incorporate findings from behavioral psychology to hook users, such as triggering, queueing,

providing variable rewards, and prompting reinvestment (C&P, 2020: 7-8). However, the

problem is worse than fostering addiction—the way that the “hooking” mechanism operates in

AE is not via a specific external object of our attention (e.g., cigarettes or slot machines) but via

the manipulation of our attention itself—whatever its objects might be. This is a significant

difference from other addictive goods and services. AE companies don’t just want people to pay

attention to a certain product (cigarettes, etc.), they want people to pay attention simpliciter.

Unlike other markets, it is not a secondary feature of AE that we relinquish control of how,

where, and why we direct our attention, but instead it is the central component of how AE

services are set up to capture our attention.

The driving motivation of AE companies is to get people to hand over as much attention

as possible, by keeping them on a highly engaging, interactive service for as long as possible;

not just to get them to buy a specific product. In AE, attention and the information it provides is

the company’s source of revenue, so the more of it, the better. In other markets, companies

would be fine economically if consumers bought their product regularly without attending to it

constantly after the purchase. Of course, a cigarette company depends on customers’ attention

returning to their products. However, there is a key difference: a cigarette company doesn’t make

more money for each additional moment a consumer attends to their product, while an AE

company does make more money for each additional moment of attention. Now, on AE services,

there is always some momentary object of our attention or other; but it doesn’t matter what the

specific object is—what matters is that we continue to pay attention to the service.

This diminished control2 over one’s attention simpliciter is deeper than the ways in which

other addictions impact one’s autonomy. To the extent that one loses control over one’s attention

(in particular, when using AE), one is less capable of fully acting as an agent—for instance, one

become less able to sustain and direct one’s focus in order to successfully achieve one’s goals.

Moreover, loss of agency—via a diminished ability to control and direct one’s

attention—is central to addiction in general. Someone who is addicted is less able to keep their

attention on other things besides their addiction, at least when they are trying to do something

else. If so, the agency-related wrong of addiction is grounded in diminished control over

attention.

Let’s further examine the difference between regular AE use versus addiction to (say)

cigarettes. While cigarettes can capture one’s attention (e.g., it may be hard to think of other

things when craving one), when the desire for it is satisfied, one’s attention is no longer diverted.

In fact, there is reason to think that cigarette smoking may sometimes increase focus on whatever

task is at hand. There are psychological effects of smoking that may enhance focus, such as

boosted energy levels. Regarding this effect, Kendall et. al (2022: 1169) note that “A large body

of literature indicates that nicotine results in an acute mood ‘boost’”. Such an energy boost could

enhance someone’s ability to focus. Moreover, Peeke and Peeke (1984) found that cigarette

smoking improved people’s recall performance, which suggests that it can improve people’s

ability to focus on the task at hand. There are times, then, when a cigarette user can regain a fair

amount of control over their attention.

On the other hand, even when one is actively using AE, the nature of this activity is such

that one’s attention is regularly diverted. Unlike cigarettes, engaging in an AE activity does not

2 In Chapter Two, I provided additional reasons why AE interferes with people’s ability to sustain attention,

including the empirical support surveyed and expanded upon by Ward et. al (2017).

lessen one’s level of distraction but actually increases it.

3 An AE user doesn’t recover their

attentional capacities while engaging with the object of their desire (an AE service). Think of

frantic, repetitive scrolling through paradigm AE services such as Twitter (X) or TikTok. Much

AE activity itself is at odds with sustained attention, so engaging with it does not solve the

problem of distraction. If so, unlike cigarettes, one cannot regain control over one’s attention

even by engaging in the desired and habitual activity. One’s attentional capacities are diminished

not just when one isn’t using an AE service but also when one is. This is a key difference

between regular AE use and addictions to some other activities and substances (e.g., focusenhancing ones), and it highlights how it interferes with autonomy in a distinct and robust way.

With other addictions one is distracted prior to obtaining the object of one’s desire (e.g., a

cigarette), but is no longer so—and maybe even more focused—once it is obtained. But in AE,

whether one is addicted or not, engaging with the object of one’s desire is itself distracting. In

this way, the extent of distraction is more pervasive for AE. When one feels a strong need to

check one’s phone (e.g., to scroll through one’s Instagram account) while doing something else,

this is distracting; but engaging with Instagram is also distracting. So, in AE, one is distracted by

a particular kind of distraction. AE users are operating in an environment that is permeated with

distraction, both when off and on AE services.

The above analysis strongly supports the claim that habitual use of (let alone addiction to)

AE is even worse—in terms of our ability to sustain and focus our attention—than addiction to

some other objects or activities. These other addictions, such as cigarette smoking, may be less

inherently distracting and may even help improve focus to some extent. The distractionpromoting environment of AE weakens autonomy by diminishing control over one’s ability to

3 Aylsworth and Castro (2021) make a related point regarding how smartphones “interfere with our ongoing…tasks”

and how this is different from addictions to cigarettes and coffee (664, cf. fn 21).

direct one’s attention. In this way, AE interferes with one’s ability to act towards one’s goals and

successfully do what one wants to do. This is especially true in domains where sustained focus is

needed, such as carrying out one’s ongoing projects and commitments.

This analysis is not meant to imply, of course, that there aren’t different dimensions along

which other addictions may be worse than habitual AE use, especially in terms of the negative

consequences for individuals. For instance, some addictions are likely much more directly

harmful to the finances or physical and mental health of individuals. (Although, there is reason to

think that AE is also implicated in mental health problems. See, e.g., C&P, 2020: 3-4.)

However, the above analysis raises doubts about whether the correct policy response to

AE is simply to regulate it like we do with other addictive products such as cigarettes (e.g., with

warning labels, as C&P suggest, 2020: 9-11). Protecting our autonomy may require a different

approach in this case.

B) Significantly More Practically Necessary, Pervasive, and Hidden

But advertising is certainly not new, so what’s wrong with AE? Someone might object

that there is a history of markets where attention is traded for services or goods, and that these

markets do not seem to be at odds with autonomy. For instance, The New York Sun, a daily paper

founded in the 19th century had a similar business model: consumers traded their attention (via

exposure to ads) to gain access to its stories, and The New York Sun made most of its revenue by

selling ad space (C&P, 2020: 2). Also, consider one-off exchanges in which someone trades their

attention to receive a good (e.g., listening to a pitch about time-shares to obtain ski tickets). So,

the objection goes, the attention economy might be just another version of this advertising-based

business model, and these types of markets are not autonomy-violating.

There are two responses available. First, one could argue that these other types of markets

and exchanges are, to some extent, autonomy-violating. Second, one could explain how AE is

significantly different from these other markets. Now, I think a case could be made that

advertising in general, especially the practice of directly trading one’s attention for goods or

services, may sometimes be autonomy-violating. However, this line of response is not necessary,

because AE does differ in significant ways from these other types of exchanges and markets.

While aspects of AE have been around to some extent before the rise of the internet and

social media, AE—the current iteration of markets in attention—substantially differs from

previous ones in terms of its practical necessity, pervasiveness, and hiddenness. First, the use of

traditional advertising markets—exemplified by the tabloid and ski ticket cases—is much less

practically necessary. Second, since these traditional markets are less pervasive, it is much easier

to obtain their goods in another way. Third, the transaction in these markets is much easier to

see—you can clearly see what is being exchanged, for how long, and in what way (e.g., two

hours of your attention for the ski ticket). These features all lie on a spectrum, and it may not be

clear when the market becomes autonomy-violating; but the cumulative effect of these features

of AE is that it is significantly different and worse than previous iterations.

AE, exemplified by services such as Facebook, YouTube, and Twitter (X), is a market

that is practically necessary and pervasive. This depends somewhat on factors such as age and

socio-economic status, but many people in the U.S. (and similar societies) own a smart phone

and use AE apps. For instance, C&P (2020: 6) point out that “97% of 12th-graders…used social

media sites, making the adoption of these platforms nearly universal.” Smartphones have also

become basically essential in order to live a normal social life—the cultural expectation, in many

aspects of social life (including work, dating, and friendship) is that one be available and

reachable on these devices. (See, C&P: 2020: 6-7.) This is part of the practical necessity of

engaging with AE. Given that people tend to be regularly on these devices and apps, the

companies and markets that trade in attention are a constant feature in people’s daily life. The

practical necessity of AE products and services for everyday needs also reinforces AE’s

pervasiveness. Basically, it’s very hard to disengage from AE and still meet one’s social needs.

For instance, it is much easier to refrain from picking up The New York Sun or trading

one’s attention for ski tickets than it is to refrain from using AE apps. While one may lose out on

the pleasure of reading an exciting news report or charging down the slopes, forgoing these

goods (which are also available via money) will not typically result in losing important social

connections or opportunities. Relatedly, one does not have to carry these attention-grabbing

items (e.g., a tabloid) in their pocket for most of the day to be connected to their social world.

Smartphones and their AE social media applications are a practically necessary and constant

presence in our lives in a way that traditional markets for attention are not.

One might object that these problems are overstated and that there are ways to get our

needs met without using AE. Couldn’t an individual just opt-out of AE and instead (say) use a

flip-phone? But this objection understates the fact that when most people regularly use

smartphones (and AE services on them), the economic and social cost of opting out makes it

much harder for any individual to do so. Consider how only a small group (2%) of people were

buying feature phones instead of smartphones in 2023.

4 This is not a large enough percentage to

count as serious competition to AE and smartphone culture.

Lichtenburg (1996) calls this type of phenomenon “technological entrenchment”, which

is when the widespread adoption of a new technology in society makes using it socially required.

4 https://www.zdnet.com/article/gen-z-are-ditching-iphones-for-100-feature-phones-and-the-numbers-dont-lie/

Once (say) email became popular, it in turn became necessary to purchase the requisite

technology and use it; otherwise, one would pay a large social cost—for example, one could

miss out on employment or social opportunities. When this point was reached, email technology

became entrenched. In general, if you need a technology to meet minimal social expectations, it

is entrenched. Adam Smith (1776) captured this point when he observed that a certain kind of

shirt was socially necessary for a day laborer in England, although it was not in other societies.

Smartphones and their AE services have become technologically entrenched. Because

almost everyone else has a smartphone, it is now the social norm that you should have one too.

Relatedly, people need a smartphone to simply remain connected with their social world. It is

difficult to use alternative technologies to fully meet one’s social expectations and needs.

Moreover, the situation with AE is like one where all the restaurants in town offer the

same menu, etc. In this case, people would be “free” to pick which restaurant to dine in, but not

the structure or composition of the menu. Our situation with AE is analogous because we can

always choose a different company (restaurant), but at the systemic level, AE’s business model

(the menu) is practically inescapable—it is the only feasible system for many important needs.

So, even if people have freedom to choose from within the existing set of AE services

(e.g., apps), they do not have a genuine choice whether to engage with AE itself. Moreover, the

fact that there are so many app choices may hide this lack of freedom.

Because there doesn’t seem to be a feasible alternative to using AE’s basic economic

framework, especially when this fact is obscured, AE seems to involve a type of domination.

Here we can draw a parallel with Anderson (2017) on the relationship between employers and

employees in many types of industries. Think of service industry jobs at (say) Walmart and its

competitors. Anderson argues that many such employees don’t have a real exit option—they can

quit their current job, but all their other “options” (jobs at other companies) are bad in similar

ways. These companies offer basically the same situation, including pay, work environment, job

expectations, treatment, etc. Even though it may appear otherwise to some people, the workers at

these companies don’t have any genuine freedom to change their situation.

In fact, without any real exit options, these private companies become like governments

for their employees. Because the workers can’t “really” leave, they must stay and either advocate

for themselves or follow the dictates of their company. Moreover, Anderson argues, since the

workers often don’t have a meaningful way to advocate for change, these companies are like

autocratic governments, because they are able to wield arbitrary power over the employees.

In a parallel fashion, the fact that we seem to have infinite choices within AE’s collection

of services obscures the fact that we have no genuine choice regarding whether to use AE itself.

It’s like having a “choice” between many different types of food, all of which fail to provide

adequate nutrition. This is a consideration about how normative conceptual frameworks can veil

reality. The way we think about freedom—namely, as a certain kind of ability to make first-order

choices—obscures the ways in which we are not free. The most effective domination is one that

is invisible. When people don’t even notice a form of domination, they are powerless to stop it.

In addition to the lack of genuine choice or say over AE and its business model, a compounding

concern is that people may not be fully aware that they are unfree at this systemic level.

AE involves another type of hiddenness—namely, it is not readily apparent to the

consumer that market exchanges are continuously taking place on AE. This problem goes

beyond the explicit display of advertisements; that market exchange is often easy enough to see.

Importantly, AE services also perpetually harvest users’ attention to generate and sell

information about them. This other aspect of AE is seamlessly integrated into the background of

people’s daily lives and habits, which makes it harder to recognize its transactional nature.

Sometimes the user may be wholly unaware that this is how AE works, or that a specific

app works this way. But even for those who, upon reflection, are aware that they are constantly

trading their attention for AE services, it is often easy to forget that this is what one is doing. The

app or website “feels free” because what one is doing on it—e.g., making plans with friends on

Facebook—seems active, agential, and unrelated to exchanging their attention for the ability to

use the app. So, there is a pervasive and false sense that most of the moments spent on one’s

smartphone apps are not market exchanges. Compared with tabloids and sales-pitches, it is much

harder to recognize the nature and extent of the market transaction. And AE companies have a

strong incentive for it to stay this way, because the demand for their services will be higher if the

services “feel free”—if these market exchanges are hidden—to their users.

Moreover, when the expectation and norm is that these services should feel free, it

becomes very hard for any other business model to compete with the attention economy model.

This incentivizes companies to further engage in these economic practices, while hiding how

they are making money from plain view.

Because of its practical necessity, pervasiveness and hiddenness, AE is significantly

different from traditional markets in attention. AE takes the logic of tabloids and sales pitches

and pushes far beyond it. App users are shown items tailored specifically to what might be

tempting for them, regardless of why they initially opened the app, on a constant basis

throughout the day, as they navigate from one relatively essential service to another. And the

exchange of attention for services has become so ubiquitous and frictionless that it is very hard to

notice it—like a fish failing to notice water. This pervasiveness, practical necessity, and

hiddenness make it significantly harder to opt-out of AE services or consent to AE transactions.

For these reasons, AE is at odds with autonomy to a much greater extent than traditional

advertising.

C) Second-order Preferences

I will now examine a related way that AE interferes with people’s autonomy. A key

agency-based wrong of AE is the likely preponderance of users who want to spend less time

engaging with AE, but who can’t—either because of habituation, addiction, practical necessity,

or some combination of these things.

There is a general sense in our society that many people want to spend less time on AEbased services. In support of this, Bursztyn et. al (2024: 1-3) found, via a large study involving

college students, that many TikTok and Instagram users would prefer for these services “not to

exist”, and that these users only use them because most other people are already on them. Given

the large network of other users, people are willing to pay to use these AE sites; however, many

users would be willing to pay for the scenario where no one could use them. Bursztyn et. al call

this type of situation a “product market trap[]” and note that it has an analogous structure to a

prisoner’s dilemma, since this group of users is stuck in a sub-optimal equilibrium (2024: 1).

Also, people presumably want to attend to the things that are valuable to them and to do

so in specific ways (e.g., a sustained and focused way). For instance, a person may want to take a

long, relaxed walk in the woods, or spend quality time with their children.

These types of desires and values generate second-order preferences (or desires)—a

preference about what one’s preferences are. At the level of first-order preferences, a user may

want to keep engaging with an AE site. However, at least upon reflection, they realize that they’d

like to spend less time on the site (or attend to other things in a more focused way), and that they

disapprove of their first order preference to keep engaging with AE. So, they also have or form a

competing higher-order preference that they not prefer to continue engaging with AE. Here, there

is a disconnect between what the agent wants, and what the agent wants to want.

Moreover, when the first-order preference to stay on an AE app wins out due to

habituation, addiction, etc., this is a case of diminished agency—because the agential (or willing)

part of a person is more closely identified with the second-order, reflective part. (Cf. Frankfurt’s

(1971) account of freedom.) We are freer as agents when what we want aligns with what we

want to want. But, in the common scenario where a person stays on an app for longer than they

wanted, it is the second-order (agential) part of the person that relinquishes control.

Of course, people are often able to eventually resist the pull of their first-order preference

to continue engaging (e.g., watching TikTok videos). But even when one is able to pull oneself

away from an AE service—when the second order preference wins out—there is still a type of

agency-based wrong: to the extent that the AE site created, enabled, and benefitted from the user

internal struggle for control, it has imposed a cost in terms of the agent’s time and energy. It

imposes a moral harm by creating conditions to be resisted, where there weren’t any before. This

is disrespectful due to its disregard for the strain it places on a person’s agential capacities, as

well as because it keeps them on the app for longer than they may have wanted. Moreover, the

pervasiveness and practical necessity of AE distinguishes it from other markets, such as the

market for sugary beverages. While this concern doesn’t apply to those who have no internal

conflicts about how they engage with AE, it is still a problem for many others.

Now, it is also possible to resist this autonomy violation by consciously trying to avoid

forming these internal conflict-inducing first order-preferences—e.g., a first-order preference for

distracted scrolling through social media. But it’s unclear how much control one really has over a

constant barrage of signals that take advantage of our psychological vulnerabilities. Moreover,

even if one can fully resist being shaped by these forces, there are psychological and social costs

to resisting the accepted way of doing things. An individual shouldn’t have to resist a default

situation that engenders distraction and shortened attention spans. Instead, the burden should be

on those who design and profit from these services, especially when practically necessary, to

change the structure that is engendering these harms. It places an unjust burden of resistance on

individuals for them to have to bear the weight of protecting their autonomy.

Of course, AE companies may start to self-regulate, as other industries (e.g., gambling)

have tried to do. For instance, smartphone companies have provided consumers with ways to

monitor their smartphone usage. However, these features—such as the option to change the color

of one’s phone screen to a less distracting gray-scale—often need to be actively sought out and

then opted into. And even if these are adopted (e.g., monitoring one’s screen time), it is unclear

whether this type of self-regulation will be enough to address the above issue, given the ubiquity,

practical necessity, hiddenness, and addictive design of AE. For one thing, AE app companies

also need to self-regulate. Moreover, since this issue has a large systemic component, beyond

simply providing resources for individuals, some amount of systemic change is likely needed.

D) Internal Preference Formation

I will now consider another autonomy concern, which is about the source of one’s

preferences. Autonomy requires that one’s own preferences come from within oneself in a

certain way. However, the attention capturing nature and design of AE interferes with

preferences being formed in this way. When engaging with AE services, one’s preferences (e.g.,

for a certain product or show) are often generated from the outside—via the promptings of a

site’s algorithm—instead of coming from within the agent. In this way, the user’s experiences

and preference formation is to some extent externally curated for them, and this is furthered by

an algorithmic feedback loop that is designed to capture their attention.

Under certain conditions, externally shaping the preference of a person is an autonomy

violation. These conditions include cases where an external force circumvents our rational

capacities to reflectively influence or endorse how our preferences are being formed. But this is

exactly what many AE products and services do. Think of a person who is shown increasingly

extreme content on sites like Facebook and Twitter: they may go down a rabbit hole and become

radicalized. Presumably, this result is often not intended by the person at any point in the

process. This is one reason their autonomy is compromised—what they want and believe, and

who they become, is not internally directed or controlled. Note that there are two problems

here—the direction in which someone’s preferences are shaped (e.g., radicalization), and the fact

that their preferences are externally shaped. I am analyzing the latter worry—the way in which

AE sites commandeer the preference formation process simpliciter. Basically, the autonomy

concern that I’m raising is, with AE, we lose sufficient control over our preference formation.

There are several related objections that one could raise to this position, and I will

consider and respond to them. In general, one might claim that preferences are sometimes

formed externally in ways that are not at odds with autonomy. As a first pass objection,

children’s preferences are largely externally controlled, such as when a parent educates their

child. In this case, the preferences formed by the child are largely out of the child’s control. This

shows that there are at least some cases where external preference formation is morally okay. If

it’s okay here, what’s the problem with AE?

This type of case is importantly different. This type of paternalism is okay for children

but unacceptable for adults (who are among those subjected to AE). It serves the child’s

autonomy (and well-being) to have a parent shape their preferences. Part of what makes the

parent-child case innocuous is that this external oversight is ultimately respectful of the child’s

basic capacities for internal preference formation and endorsement—once they are well-equipped

to exercise them. In Book IX of Republic, Plato makes the related point that we should only let

our children govern themselves after we give them the appropriate character or constitution.

A related reason why we think that paternalism is okay for children is that a parent

typically has their child’s best interests at heart. A good parent will want to instill preferences in

their child that will enable the child to live well. And until a certain age, we think that a person is

not yet fully equipped to govern themselves or independently form good preferences. Preference

formation will happen regardless of whether one intervenes or not, and so children require care

and external constraints.

But AE services do not aim at shaping preferences that are in one’s best interest. Instead,

the primary profit incentive is to capture as much attention as possible, regardless of whether the

preferences and habits formed are in a person’s best interest. This should raise some serious red

flags about AE use in children—it interferes with their current autonomy (preference formation)

without the corresponding benefit that parents aim to provide, namely promoting their well-being

and future autonomy. Moreover, parents are losing some control over this process. This type of

concern helps to explain the recent backlash to AE services, such as Meta’s Instagram app, given

the undue amount of influence they have over teenagers’ preferences. In sum, AE is different in

key respects from the parent-child case: many of its users are adults, and even for children, AE

companies don’t have the same mitigating beneficent intentions that parents possess.

Consider another objection—one might argue that adult’s preferences are always formed

by their environment (including by other people), at least to some degree, so what’s the problem

with AE? For instance, in previous advertising markets, we might form preferences for (say) a

vacation destination that we did not have before viewing the highway billboard or TV

commercial. Moreover, as in these other cases, what you are shown by AE sites are just

suggestions that (unlike subliminal images) are open for you to rationally evaluate and decide

whether or not you want to endorse them.

In response, I first draw upon my earlier argument that AE is much more pervasive,

practically necessary, and hidden than previous advertising environments. As argued above,

using many AE services has become a social necessity. So, it is very hard to opt-out of AE, and

hence the preference shaping process. People could choose not to watch TV or pay attention to

traditional ads (e.g., in a magazine), but it is much harder to ignore AE’s social media and news

feeds. Relatedly, to the extent that people are addicted to an AE service, they have some amount

of weakened agency, in that they have less (second-order) control over their decisions in matters

regarding their addiction. So, what initially appeared to be suggestions (e.g., about which

programs to watch) are no longer mere suggestions.

This also applies to a related objection: namely, that one can consent to being subject to

external preference shaping by agreeing to use AE—that this mitigates or obviates the autonomy

violation. But the above considerations (e.g., practical necessity) makes it questionable whether

the decision to use AE services counts as legitimate consent. Also, while this is not strictly

needed, perhaps one cannot agree to give up such a basic part of one’s autonomy (some limited

control over preference formation), at least at the level required by regular or habitual AE use.

The high level of personalization of AE’s “suggestions” are also part of the problem.

Unlike traditional advertising, people who are simultaneously using the same AE site may be

shown vastly different content and ads. This level of fine-grained preference suggestions is much

more paternalistic than the generic suggestions offered by previous advertising markets. Like a

parent who knows their child well (but without the same beneficent intentions), AE’s algorithms

are more able than earlier advertising to respond to detailed info about a user, in order to mold

their preferences in one direction or another.

Now, it is possible that sometimes the algorithms behind the recommendation process are

merely responding to pre-existing preferences rather than shaping preferences. This would be

less concerning, but the pervasiveness (and hiddenness) of this process seems to increase the

likelihood that people’s preferences are being shaped at least some of the time.

Continuing along these lines, the objection about external preference formation

downplays the extent to which the process behind what people are shown, and how it is

displayed, are hidden from conscious recognition. This is another facet of AE’s hiddenness—the

ways in which the algorithm works at generating recommended content is opaque to us. But the

process by which these suggestions are generated seems relevant to whether we should rationally

endorse how our preferences are formed by them. If we have no control or access to the

recommendation process, we won’t know (say) whether and why it is leading us in a particular

direction. Having such knowledge seems relevant for informed choice and evaluation. This

makes it harder to say that the preference formation is rationally guided from within the agent.

Of course, other people are also part of our environment, and they play a role in shaping

our preferences. There is also some difficulty in determining another person’s recommendation

processes and their intentions behind it. But this doesn’t seem objectionable in terms of

autonomy, so what is so bad about AE’s hiddenness? However, this type of hard-to-glean

environmental influence is relevantly different from AE. When a person makes a

recommendation to you, it’s easier to determine the thought process behind it as well as their

intentions. We have significantly more practice and knowledge regarding how to make such

determinations than we do with algorithms (or AI). It seems to be in principle more accessible to

us. So, the hiddenness problem is much harder to resolve for AE.

III. AE and Competing Obligations

Castro and Aylsworth (2021) have argued that individuals have a duty to minimize their

AE usage in order to safeguard their autonomy. In the previous section, I provided a distinct and

multi-pronged argument as to why AE is deeply at odds with our autonomy. My above analysis

strengthens whatever case there may have already been for a moral obligation to curtail one’s AE

use in order to protect one’s autonomy.

Simply being conscious of the autonomy-based pitfalls of AE isn’t enough to avoid such

an obligation; for one thing, regular use of AE—even if one is aware of it—still impedes one’s

control over the ability to sustain and direct one’s attention. While targeted legislation might help

lessen this negative impact, there is a difficult collective action problem involving strong

opposing interests that must first be overcome (cf. Chapter One). An obligation to minimize

one’s AE use will remain in place until successful legislation is implemented. Moreover, the

considerations from my other chapters add to the moral weight of this obligation; these include

my arguments that AE is at odds with important virtues and attitudes about attention.

However, there is an important caveat: the obligation to disengage from AE applies to the

extent that this is possible. Due to considerations like AE’s pervasiveness and social necessity,

not everyone will be able to fully do so. Consider the broader question of whether it is morally

permissible to drop out of society, at any level of development. In Walden, Thoreau (1854)

describes how he was able to disengage from everyday life. But dropping out completely, like

Thoreau did for two years, is an extreme case. Both in general, and when it comes to AE, this

level of disengagement may only be available to a privileged few. Moreover, some people might

not even be in a position to substantially limit their AE use, let alone completely disengage from

it. To the extent to which this holds, it would diminish, or perhaps even negate, their obligation

to disengage from AE. Competing obligations to oneself or others may outweigh it.

In light of this, I’d like to raise and examine a connected moral complication that is

engendered by AE. Namely, the obligation to protect one’s autonomy may be in tension with the

obligation to adequately attend to our larger social and political world. I will explicate why AE

makes it harder to adequately fulfill both obligations, and I will offer some suggestions for what

might be needed to counteract this.

AE places people in a moral predicament about whether to use it. On the one hand, we

arguably have at least a pro tanto moral and civic obligation to stay reasonably informed about

important political issues. Also, we certainly have some social obligations to help and attend to

others, especially our friends and family; and due to AE’s pervasiveness, we might be less able

to do so adequately if we refrain from AE activities. However, on the other hand, out of respect

for our autonomy, or out of regard for the value of sustained attention, we also have an

obligation to ourselves to limit our use of AE or disengage from it.

Consider the obligation to stay reasonably informed about larger social and political

issues.5 To the extent that AE pervades many of the typical ways of obtaining information about

5 I will focus below on this obligation. However, I believe there are important parallels between this case and the

obligation to adequately attend to one’s social network, especially one’s friends and family.

important issues, we seem to be in a moral bind. There is reason to think that it is hard to escape

AE while staying informed: this difficulty applies not just to sources of news such as social

media, but also to online and cable outlets. Social media is obviously part of AE, but other

sources must compete in the AE environment. For instance, traditional news sources have a

strong incentive to have their articles shared on social media, given its massive user base. There

is a limited amount of attention (in the sense of clicks, likes, etc.), and with the advertising-based

(and even subscription-based) revenue model of news sources, this will engender intense

competition for attention. So, these impacts will expand beyond core AE services like social

media apps to more traditional news sources.

The moral issues surrounding AE, including how it thwarts our autonomy, apply to many

common types of news sources. For instance, when a news source’s primary goal is to maximize

attention, we may be exposed to very upsetting content that is designed to stir up an intense

emotional response and capture our attention.

6 And if one wants to object that news has always

been like this, I argued above that AE is significantly different in terms of its ubiquity—the

volume and degree of this practice of has been substantially turned up by AE. For example, to

even read through an article on many traditional news sources’ websites, a person must often sift

through an interspersed series of emotionally laden news headlines.

It is worth noting that this problem lies on a continuum, and that some news sites,

including for-profit ones, do a better job reporting the news in a way that is less autonomyviolating; as opposed to distracting people, influencing their preferences, etc. But AE’s model

incentivizes these autonomy violations, and its pervasiveness shifts things in that direction.

Either way, scrolling through social media to obtain and read news just exacerbates these issues.

6 Epistemic issues, such as misinformation, may generate an additional obligation to limit or disengage from AE.

So, AE places us in a moral predicament because it makes paying attention to important

social and political issues at odds with other important moral obligations. That is, to protect our

autonomy from AE, we may have to—to some extent—fail at our obligation to attend adequately

to important issues.

Of course, one might question whether we do have a pro tanto moral or civic obligation

to stay informed about social or political issues. However, there is reason to think that, with the

amount of information available to us via the internet, we might have more of an obligation than

we otherwise would. With the internet, especially social media and AE, people are inundated

daily with many different things, causes, events, etc. worth caring about. (cf. Burkeman, 2023.)

Our attention is often brought to whatever things are most tragic in our regional or national

environment. This increases our exposure to more items that are not only unworthy but also

worthy of our engagement. An increased exposure to things worth caring about, in turn, increases

our democratic obligation as citizens to at least pay attention if not act.

But companies are relying upon this sense of obligation, among other things, to capture

more of our attention. In fact, in this way, they are instrumentalizing our attention-based

obligation—to stay informed and involved—as a means to making more profit. So, for this

reason as well, AE engenders a dilemma of either protecting ourselves from such practices or

fulfilling our obligations to others: that is, we can either protect our autonomy (by not paying

enough attention) or remain adequately informed (but not both).

Now, one might object that there are feasible ways to remain informed without engaging

with AE. For instance, PBS news is still available, and this doesn’t follow AE model. Also,

people can still subscribe to the paper version of traditional news sources like The New York

Times or The Wall Street Journal, which (though they may contain attention grabbing items like

vivid front-page stories) seem generally further removed from the attentional milieu of AE.

In response, part of staying adequately informed about one’s society is that one needs to

be aware of the channels and ways in which other people are getting their information. This, in

turn, generates an obligation to attend to these sources of information. Since many people,

particularly younger generations, are getting their news from social media—or at least from

online news sources and smartphone apps—this means that one ought to maintain a certain level

of engagement with AE-based news sources. Otherwise, one would remain uninformed about a

significant part of society’s information landscape.

Moreover, because many people in our society get their news from AE-based sources, it

becomes harder to avoid its influence. This includes the way AE tends to steer people’s attention

to sensationalized and upsetting content. By influencing those around us, AE degrades part of

our attentional and informational ecosystem, even if we are deliberately attempting to disengage

from AE. Like pollution, others’ consistent use of AE news sources can result in externalities, in

this case on our own attentional capacities.

7 If so, it may be harder to disengage from AE, while

remaining adequately informed, than the objection supposes.

Let me introduce a final complication, namely, that there is not only an obligation to

attend to injustices, but also to attend to them in certain ways. For instance, attending to

injustices in an amused, let alone distracted, way is not appropriate. Instead, the way we attend to

these things arguably should be in a focused and conscientious manner. But AE interferes with

this as well because, as a way of maximizing attention capture, it tends to prompt users to move

7 Perhaps one’s individual contribution to a degraded attentional and informational environment supports an

obligation to others that one should minimize one’s own AE use. If so, this would be a complementary duty to

Aylsworth and Castro’s (2022: 2) position, namely, that an individual has a “general…imprecise” duty to others to

help them minimize their AE use.

from one tragic story to the next. AE has significantly exacerbated an issue that concerned

Kierkegaard (1846: 35), namely, that his era was one of “advertisement and publicity” combined

with complete inaction. In particular, each injustice does not receive the undivided and careful

attention that it deserves. This connects to the way AE interferes with commitment. (cf. Chapters

One and Two.) It is hard to maintain focus on any one issue, so none of them end up getting the

proper focus that is needed to fully engage with them.

So, even if people are attending to injustices via AE services, they may not be attending

to them in the right way. Since our attention is stretched thin by AE, it becomes harder for each

of the many important issues in our social world to get the attention that they deserve, in the

manner they deserve. This final point makes the tension between AE and our attentional

obligations even starker. It’s not just that AE makes it harder to fulfill obligations to protect our

autonomy or cultivate good attentional habits, but that even if we attempt to pay attention to our

larger social world, AE interferes with our ability to attend to it in a proper way. Our

predicament is that we should engage with the world but that the ubiquity of AE degrades our

ways of interacting with the world (e.g., our attentional habits), as well as our autonomy.

The above analysis provides strong reason for preserving and expanding alternative ways

of engaging with the world. Informational content about the world can come in different forms.

Some of these outlets (e.g., certain social media apps) tend to be significantly more distracting

and preference shaping than others (e.g., PBS and other reputable news sources), and one should

both be aware of this and have the option to choose judiciously between them. Content produced

by entities that are respectful of people’s autonomy (e.g., that present the news in a way that isn’t

distracting and diverting) ought to remain available and be widely promoted—it may help

alleviate the moral bind of these competing obligations to some extent. But alternative models to

AE are at risk of losing revenue and funding; and this would make it harder to obtain content

outside of AE. If so, people may have a moral obligation to support these types of outlets.

IV. Conclusion

In this essay, I argued that AE violates our autonomy in multiple ways. These ways

include interfering with our basic capacity to direct and sustain our attention; AE’s

pervasiveness, practical necessity and hiddenness; and that AE inhibits our ability to satisfy

second-order preferences, as well as internally form our own preferences. Not only are these

violations problematic in themselves, but diminished autonomy regarding one’s attention

negatively impacts one’s ability to live well. (I explore the connection between flourishing and

AE more directly in other chapters.)

I also examined an implication of this argument: namely, that if AE violates our

autonomy, we have at least a pro tanto obligation to protect ourselves from AE; but that this

conflicts with an obligation to adequately attend to and engage with the world (and to do so in

certain ways). I argued that this dilemma is hard to resolve because of how pervasive AE is and

the manner of attention it promotes in its users. One potential remedy is the development and

promotion of non-AE sources of news and social connections.

REFERENCES

Aylsworth, Timothy and Clinton Castro. 2021. “Is There a Duty to Be a Digital Minimalist?”

Journal of Applied Philosophy, 38, no. 4: 662-673.

Aylsworth, Timothy and Clinton Castro. 2022. “On the Duty to Be an Attention Ecologist”.

Philosophy & Technology. 35 (13): 1-22.

Anderson, Elizabeth. 2017. Private Government: How Employers Rule our Lives (and Why We

Don’t Talk about it). Princeton University Press.

Bhargava, Vikram and Manuel Velasquez. 2021. “Ethics of the Attention Economy: The

Problem of Social Media Addiction,” Business Ethics Quarterly 31(3): 321-359.

Burkeman, Oliver. 2023. Four Thousand Weeks: Time Management for Mortals. Picador.

Bursztyn, Leonardo, Benjamin R. Handel, Rafael Jimenez, and Christopher Roth. 2024. “When

Product Markets Become Collective Traps: The Case of Social Media”. NBER, Working Paper

No. 31771.

Castro, Clinton and Adam Pham. 2020. “Is the Attention Economy Noxious?” Philosophers’

Imprint, 20, no. 17: 1-13.

Frankfurt, Harry. 1971. “Freedom of the will and the concept of a person”. Journal of

Philosophy, 68 (1): 5-20.

Kendall, Ashley, Donald Hedeker, Kathleen R Diviak, and Robin J Mermelstein. 2022. “The

Mood Boost from Tobacco Cigarettes is More Erratic with the Additions of Cannabis and

Alcohol”. Nicotine & Tobacco Research, 24 (8): 1169–1176.

Kierkegaard, Soren. [1846]. The Present Age. Translated by Alexander Dru. Harper Torchbooks,

1962.

Lichtenberg, Judith. 1996. “Consuming Because Others Consume”. Social Theory and Practice

22 (3): 273–297.

Nozick, Robert. 1974. Anarchy, State, and Utopia. Basic Books.

Peeke, S.C. and Peeke, H.V.S. 1984. “Attention, memory, and cigarette smoking".

Psychopharmacology, 84, 205–216.

Plato. Republic. Translated by C.D.C. Reeve. Hackett, 2004.

Sandel, Michael. 2012. What Money Can’t Buy: The Moral Limits of Markets. Farrar, Straus, and

Giroux.

100

Smith, Adam. [1776]. An Inquiry into The Nature and Causes of the Wealth of Nations.

Clarendon Press, 1976.

Thoreau, Henry D. [1854]. Walden and Other Writings. Bantam Books, 1962.

Ward, A. F., Duke, K., Gneezy, A., and Bos, M. W. 2017. “Brain Drain: The Mere Presence of

One’s Own Smartphone Reduces Available Cognitive Capacity”. Journal of the Association for

Consumer Research, 2: 140–54.

101

CHAPTER FOUR: CORRUPTION AND THE ATTENTION ECONOMY

I. Introduction

The attention economy (AE) is a market that is characterized by two types of transactions:

[1] people exchange their attention for access to a new media service or product (e.g.,

Instagram); [2] these companies then sell our attention, and the information it generates, to

advertisers. (Following Castro and Pham, 2020:2.) As I will expand upon below, in the first type

of market transaction, people use their attention as currency (a widespread means of exchange)

to access AE services. In the second type of transaction, our attention is the companies’ input

resource and product—they generate revenue by capturing and selling our attention.

While it may initially seem innocuous to exchange attention in a market for access to

goods and services, we have now seen that AE has disruptively expanded into a large swath of

our economic and social lives. Particularly, in Chapters Two and Three, I argued that AE

significantly interferes with sustained attention, and that it is much more practically necessary,

pervasive, and hidden than previous markets in attention. These types of issues have led us to ask

whether and how this market is morally problematic. In this chapter, I will examine AE in detail

through the lens of one type of objection to markets, namely, the corruption objection.

The corruption objection to markets in particular goods is that such markets (for certain

goods) would “express and promote” the wrong attitude or valuation of these goods (Sandel,

2014: 10, my emphasis). For instance, some argue that a market in sexual services would send

the wrong signal and promote bad attitudes about the value or purpose of sex. (cf. Anderson,

1993.) This worry about corruption is distinct from worries about (say) fairness or autonomy.

102

In this chapter, I examine whether AE is subject to the corruption objection, and I argue

that it is: viz, that AE expresses and promotes corrupted (or degraded) attitudes and values about

our attention. In section II, I argue that a) AE expresses and promotes certain attitudes about

attention—that it is instrumentally valuable, commensurable with other market goods, valuable

only for its quantity (and not its quality); and that b) these attitudes about attention are morally

degraded from multiple standpoints, with a focus on the virtue ethics perspective. To bolster my

case in this section, I argue that the scope of AE and the transactions that constitute it is

extensive, going well beyond visible advertising. Then, in section III, I defend my position about

AE against Brennan and Jaworski’s (2015) prominent criticism of symbolic objections to

markets, namely, that these objections fail due to the contingent meaning of markets.

II. Attention and the Corruption Objection

A) The Meanings Expressed and Promoted by AE

An initial way to examine whether a market in a particular good is objectionable on

corruption grounds is to ask whether it falls into one of two categories: is it the type of good that

can be properly traded for regular goods and services, or is it disrespectful or improper (in terms

of its meaning) to use it in this way? Paradigm cases of the former category include toasters, cars

(and, of course, money); a paradigm case of the latter are sexual and reproductive services. Now,

there is some debate about whether there are goods that fit into this second category. For the

moment, I presume that there are such goods. Later, with an eye to AE and attention, I will rebut

Brennan and Jaworski’s (B&J) prominent argument that there are no such goods.

So, the question becomes, where does attention fall—is it more like cars and other

consumer goods or is it more like sex? At first glance, it appears that it is not objectionable to

103

trade one’s attention in a market transaction, e.g., for access to information or entertainment.

Consider someone who (say) trades their attention, by listening to a two-hour sales pitch about

condos, in order to access a ski ticket. Attention seems to be unlike sex in that it is not inherently

improper or demeaning to trade it for other consumer goods and services. We might wonder

whether the mere act of exchanging, say, a few minutes of one’s attention for access to a

television program, as people do for broadcast TV, expresses or promotes a degraded attitude

about attention. Does this mean that attention falls into the other, non-objectionable, category?

I don’t think that this is the end of the discussion. I think that there is a key distinction

that hasn’t been clearly or fully articulated in the debate about markets and corruption. So far,

within the context of the corruption objection, we’ve carved up goods in a binary way (i.e., those

that are either proper to markets or improper), but I think that there is a third category. Namely,

certain goods are morally permissible to use as currency in individual exchanges but morally

should not be a primary currency (nor a central commodity) for a large-scale market or

economy. When a good is a primary currency, it is the main means of exchange for a variety of

different goods and services. The obvious example of a primary currency is money, but other

goods could also play this role. For instance, we can imagine an agrarian economy where the

primary means of exchange is corn (or IOUs for a certain amount of corn).

Now, there is reason to think that some meanings are expressed by any market

exchange—namely, that the goods in question are commensurable with other market goods. For

two people to exchange a good for money, this seems to imply that the good involved is treated

as on the same scale of value as money. This is because money as the means of exchange

connects the value of that particular good to any other good that can be exchanged for that

amount of money. This seems to express that the value of that particular good is on the same

quantitative scale of value (i.e., commensurable) with all the other things money can be used to

104

buy. Goods that are regularly traded for money are treated as having the same type of

quantitative value.

For even more reason, this meaning—commensurability—is expressed when a good is

used as a primary currency for an entire market. This is because the good used for currency is not

just one item on the market, but a primary vehicle of exchange.

I think that attention is this third type of good. Although there may be a slight air of

disrespect or impropriety to trade attention in an individual exchange (e.g., attention for a ski

ticket), this may not rise to the level of moral wrongness. However, when it is the main means by

which people are accessing a wide range of goods and services (as well as a central commodity

for an economy), then it crosses a threshold—it then expresses and promotes bad attitudes about

what attention is and why it is valuable. This is because when something becomes a primary

currency for a large market or economy, it is unequivocally being used and treated as a means to

obtain many other goods and services in a large segment of the economy. So, at this point, it is

commensurable (on the same scale of value) with any other market good, and its social meaning

becomes similar to money—that is, primarily instrumentally valuable for use in exchange.

In particular, attention is treated this way by AE. That is, AE expresses and promotes an

attitude that attention is commensurable with market goods (e.g., with cars, toasters, etc.), as well

as instrumentally valuable, because it is a primary means of exchange for accessing AE services.

At this stage, one might be wondering how we can determine whether some good is a

primary currency in a given economy, and whether, in particular, attention is a primary currency

in AE. I admit that it is tough to explicitly delineate the criteria for when this is the case. What I

want to suggest, however, is that there is plausibly a threshold level at which a good becomes a

primary currency. Moreover, given the complexity of large-scale social phenomena, it is likely a

vague concept—in the sense that the boundaries to what phenomena it captures are not sharply

105

defined or definable. So, it is not easy to specify exactly when a good, such as attention, becomes

a primary currency of a given market or economy. However, like the sorites puzzle—where (say)

at some point a collection of sand grains becomes a mound of sand—at a certain point attention

passes the threshold into being a primary currency.

Let me say more about what features might help indicate that the threshold has been

crossed for attention in AE. For starters, in Chapter Three, I argued that AE is pervasive and

practically needed, especially for important social and work-related activities, such as connecting

with friends and colleagues; the pervasiveness and practical need to use AE reinforce each other.

These points provide reason to think that the scope of AE covers a large swath of economic

activity. Due to its ubiquity in our social lives, people repeatedly interact with AE services like

social media apps on a daily basis.

Next, when it comes to AE products, accessing them often does not require a monetary

fee. In the marketplace for smartphone apps and websites, many of them (e.g., Instagram or

TikTok) are “free” in this sense. Since these products are often being provided by for-profit

companies, there must be some way for the companies to generate revenue.

The means of exchange (for access) in these types of “free” apps is user attention, which

is then used to generate revenue in multiple ways. First, these companies sell user attention to

advertisers and, in turn, users are shown traditional ads. This transaction is easy enough to see:

people gain access to AE sites in exchange for viewing ads. Second, there is product placement,

where products are woven into the user experience more surreptitiously—e.g., a social media

influencer happens to be wearing a certain brand of clothing in a given post, and they and the site

make money based on (say) the total number of views. In both these cases, the users’ attention is

itself the company’s product.

106

Importantly, there is a third way that attention is exchanged for access—namely, many

AE companies are harvesting people’s attention, using this as an input resource, to generate

information about their preferences, etc. How is attention currency in this case? The revenue that

is generated by selling this data to third parties, using attention as a raw material, allows AE sites

to not charge a monetary fee. (Moreover, even if an AE company isn’t selling this type of

information to third parties, they may still use this information to maintain and enhance their user

base, which is a source of additional revenue.) If these AE companies didn’t engage in this

practice, then, in order to recoup the lost revenue, users would have to pay some amount money

for access to make up the difference. Exchanging one’s attention (as an input resource for the

company) to reduce the monetary cost of access is to use one’s attention as currency instead for

accessing these sites.

To the extent that these ways of accessing services and generating revenue are a

widespread practice across AE, attention plausibly has passed the threshold into being a primary

currency. How widespread must it be? The answer again will not be in precise percentages, but

instead in a general sense that exchanging attention for products and services is the norm when it

comes to a given domain. I think it is quite plausible to claim that this is the case in the domain

of apps and websites (i.e., AE).

B) The Extensive Scope of AE

Now, someone might object that there are new media companies that charge people

directly to use their service or product. There are sites, like Hulu, that charge a monthly

subscription fee. Do they and their subscribers also treat or use attention as currency? Are these

companies still part of AE? The objection here is that perhaps a significant part of the new media

107

economy is based on money instead of attention. We might well wonder how widespread the use

of attention as currency actually is.

In brief, my answer is that attention is generally used or treated as currency, as well as a

commodity, even by the types of new media companies that charge users monetary fees. I will

defend the claim that these companies are part of AE—that their users exchange attention for

services and that they generate revenue both directly and indirectly (as a resource) from people’s

attention.

Let’s take Hulu as an example. Importantly, Hulu engages in the practice of product

placement—in many shows, products, such as different beverages, are made conspicuously

salient to viewers.1 In this way, Hulu is making money directly off our attention, because it sells

user attention to the relevant advertisers. All this provides reason why more attention equals

more money for Hulu (and similar subscription-based sites), because if there are more views,

then companies will pay more to have their products incorporated into Hulu’s shows. Product

placement revenue gives Hulu a clear incentive to increase the amount of time that people spend

on their site, even though they have already paid their monthly subscription fee. Hulu can make

more money either by gaining additional subscribers or by capturing more attention from its

current users (or both).

This is reason to think that streaming companies like Hulu do not follow a pure

subscription model. Instead, in addition to charging a monthly fee for use, they are still regularly

commodifying our attention in other ways. One of these ways is product placement, which

provides an incentive to quantify, monitor and increase viewing hours. By enabling Hulu to

make money via their attention, users pay less than they otherwise would for access to its

services. Because there is a gap between the full price of access (what it would cost without

1 https://www.businessinsider.com/hulu-exec-work-brands-product-placement-differs-netflix-rivals-2019-10?op=1

108

attention exchanges) and the actual price paid, this constitutes a transaction of attention for

services. To do so is to treat attention as currency.

Moreover, advertisers and other businesses generally want to know what Hulu’s (and

other subscription-based sites) users are paying attention to—this large quantity of data

constitutes another potential revenue stream for Hulu and other new media companies. Once

collected, this information can be sold to them. Moreover, it appears that Hulu engages in such

practices.2 This generates an incentive to keep users’ attention for longer in order to exchange it

for additional revenue. While this exchange of attention for money is less direct, it is still

essentially about capturing attention so as to sell the information it provides to other parties. In

this case, increased attention (as an input resource), and the information that this attention

generates, is a revenue source for companies. And this lowers the monetary cost of access for its

subscribers, who again are trading their attention in order to get this lower price.

The above analysis bolsters the argument that AE’s market of websites and smartphone

apps treat attention as a primary form of currency, because even sites that charge users monetary

fees (e.g., subscriptions) still make some of their money directly off user attention in the same

ways (including advertising) as pure AE model companies. This enables them to charge users a

lower monetary fee in exchange for their users’ attention. In this way, I am broadening the scope

of what types of companies use attention as a primary currency to include (say) Hulu, which has

a subscription model. I argued that they are making money in other ways that go beyond

subscriptions. Importantly, these ways include capturing attention and selling it to third-parties.

Moreover, to the extent that people are less aware of this aspect of the new media

economy, it is a hidden problem—even when we pay with money, we are still also often trading

our attention for services. The full price of the service is higher than the money we exchange for

2 https://press.hulu.com/privacy-policy/#SharingWithThirdParties

109

it—our attention is used as additional currency in order for us to access the service. In fact, this

might be even more insidious, because the transactional nature of exchanging our attention for

services is more hidden with (say) subscription-based companies like Hulu. We may think that

by paying a subscription fee, we are opting-out of the AE model. In this case, it might be less

clear to users that their attention is part of the economic transaction, and this lack of transparency

raises autonomy concerns.

Importantly, my analysis of the scope of AE adds to the concerns raised in my other

chapters regarding AE’s negative relationship with autonomy and flourishing. This is because it

shows that the scope of AE is larger than perhaps initially thought—it is not limited to products

and services that involve pure AE models (that only trade services for attention—e.g., Instagram)

but also extends to those that use mixed models as well (e.g., Hulu).

For these reasons, the attention capturing nature of AE products and services, and the

moral issues that correspond to them, extend to a much larger domain. To take a key example,

subscription-based models can be (and often are) still part of AE. In particular, that means that

attention is a primary currency for an even larger domain than one might have initially thought.

This matters because the pervasiveness of AE, with its use of attention as currency, increases the

magnitude of the expressive wrong it communicates and its degrading impact on our attitudes

and values. The attitudes and values expressed by treating attention in this way—as instrumental

and commensurable—permeate the tech landscape of smartphone apps and social media.

C) Attention and AE: Corruption Concerns

I argued above that attention is treated as currency within AE; that AE constitutes a broad

sphere, which includes many subscription-based companies; and that this expresses and

promotes certain attitudes and values about attention (e.g., as instrumentally valuable).

110

Somewhat implicit so far has been the further claim that this is morally problematic—that these

are bad attitudes and values to express and promote. For any corruption objection to go through,

it requires that there are better ways than this to value certain goods, such as our attention. (cf.

Sandel, 2012) Now, I think it’s intuitively plausible that there are better ways to value our

attention than those expressed by treating it as currency, and that it is thus problematic to treat it

in this way. That being said, let’s examine some reasons in favor of this claim.

For it to be morally wrong to treat attention primarily as currency—as instrumentally

valuable, primarily for use in exchange, commensurable, etc.—it would have to be an improper,

disrespectful, or harmful way to think of and value attention. Here is an argument for why this is

indeed the case. Attention is a core part of a person—in particular, it helps comprise their

identity and autonomy, and (relatedly) it plays a key role in habit and preference formation. I

think we have strong reason to think that, to a significant extent, we are what we attend to and

how we attend to it. There is a close connection between what we attend to, how we attend to it,

and what we value. Typically, the more one values something, the more (and more carefully) one

attends to it, and vice versa. This is part of how our attention helps to shape our identity—via its

impact on our values, preferences, and the habits we form.

Basically, the idea is that people are (in part) what they attend to and vice versa.

Attention doesn’t merely belong to a person, but instead it is a key, integral part of them. If it

were diminished, commandeered, or removed, unlike (say) a finger, one would lose one’s

integrity as a person. In this way, it is a necessary constituent of who someone is. What we

attend to (and how we attend to it) expresses who we are, and it helps to shape and maintain our

identity, including our values and commitments.

Given this, thinking of attention as commensurable and instrumentally valuable is indeed

morally incorrect. This is because our attention is a constitutive part of our core person. To treat

111

people, including ourselves, in this way would be to think of and value them incorrectly. Let’s

look at this from a deontological perspective. From this perspective, it treats a core aspect of who

a person is as a mere means for exchange. To treat and think of this essential aspect of a person

in this way is disrespectful to their dignity as persons, because people have intrinsic value and

are not commensurable with market goods.

Moreover, as argued in Chapter Three, being able to sustain and direct our own attention

is a key part of being an autonomous agent. Without this capacity, for instance, we would

struggle to internally form our own preferences, fulfill long-term commitments, or even stave off

addiction. But expressing and promoting the view that attention is like currency—and can be

bundled up into pieces and sold without any loss of value—ignores the importance of sustained

attention for autonomy; in this way as well, then, AE is disrespectful to the people using it.

I will now examine the virtue ethics perspective. AE’s view of attention (as instrumental

and commensurable) expresses and promotes the idea that attention is valuable in this way; but,

as argued in Chapter Two, attention is a constitutive part of many aspects of flourishing—e.g.

self-knowledge and committed relationships. Here I can fold in all the reasons from this chapter

as to why AE is deeply at odds with flourishing. By expressing and promoting this way of

viewing attention (e.g., downplaying its intrinsic importance), AE encourages attention-based

habits and character traits that interfere with our ability to live well. Relatedly, thinking of

attention as something that can (without loss) be diverted, spread, and exchanged away is in

tension with proper ways of attending to ourselves, others, and our environment; instead, we

need to develop and implement other good attentional habits, especially sustained attention,

mind-wandering, and flow. This is the corruption worry from a virtue ethics perspective—

namely, if we are concerned with living well, then AE expresses and promotes a bad way of

valuing our attention.

112

Let’s expand and add to the above. By using attention as currency and as a product,

especially in such a sizable market, AE expresses and promotes the idea that attention is a

discrete means to accessing services and making money. By placing it in this role, attention is

treated and thought of as piecemeal by individuals and companies, in addition to being treated

instrumentally. But, under a different way of thinking, attention has a holistic component: that

is, it is valuable for the coherent narrative it can help us form about our lives, and essential to the

integrity of an individual. The coherence and integrity that a person’s sustained, undivided

attention provides may be part of a fully flourishing life. However, this view about attention is at

odds with the piecemeal, discrete, instrumental view of attention expressed and promoted by AE.

Relatedly, because it treats it as currency, AE expresses and promotes the idea that our

attention is valuable solely for its quantity. However, there is a distinction between quantity of

attention and quality of attention. AE sites make money by quantifying our attention via

measurable proxies, including the number of ad clicks, the amount of time people spend on their

site, etc. This is a particular case of how markets focus on what’s measurable. Moreover, using

attention as primary currency treats it as commensurable with other goods.

AE is all about quantification of attention. This collapses and obscures potential

qualitative differences between different ways of attending to the world. Instead, it treats all

forms of attention as being on the same scale of value. When quantification is what matters, a

certain amount of frantic attention paid to (say) false, polarizing, or emotionally upsetting

content is just as valuable as that same amount being calmly and reflectively paid to (say)

educational content. (Note that the idiom “paying attention” also reflects and promotes a marketlike view of attention.) But there are better and worse ways to attend to ourselves and the world.

Even if one disagrees with this last claim, I note that doing so is to take a moral

position—i.e., that there are no better or worse ways to attend to things. But by focusing on

113

quantity of attention without regard to quality, a widespread AE market presumes, reflects, and

promotes that position. This is another aspect of the corruption objection: AE market expresses

and promotes a way of viewing and valuing attention—as quantifiable and valuable as such. To

this extent, AE market is not morally neutral.

Now, just as we’ve inquired into the value of attention, we can similarly ask whether

there are certain purposes for attention, whether natural or socially constructed. By treating it as

currency, AE promotes the view that attention is primarily for private ownership and use, as a

means that can be exchanged without loss for other market goods. But if attention is primarily

for other purposes, it would alienate people from such purposes to use and treat attention in this

way (as currency). There is a case to be made that attention does indeed have other key purposes,

including forming part of our integrity and character as discussed above. (Note that to claim that

attention has no purpose, or that none is better than another, is a moral view that needs defense.)

Consider briefly the Marxian view that marketizing and commodifying labor alienated

people from its purpose. Marx argued that labor is supposed to be intrinsically valuable, creative,

social, and expressive of our core selves, but that capitalist labor markets, by treating labor as an

exchangeable means of production, turned it into an instrumental and private good—thus

alienating people from labor’s purpose (and themselves). (See Wolff, 2002.)

I’d like to suggest an analogy to attention here. If attention is central to who we are, and

if it is better thought of as intrinsically and holistically valuable, then it is alienating to quantify

it, break it up, and exchange it as a means for other goods. By using our attention as currency and

as a product, we are using it (and ourselves) for a purpose that it is neither primarily nor properly

for—if so, this alienates us from our attention and ourselves.

An objection to Marx’s view about human nature and labor is that it is not convincing, as

it relies on a somewhat idiosyncratic view of the purpose of labor and its centrality in human

114

flourishing. If Marx is wrong about the purpose of labor, so much the worse for his claim that

labor within capitalist systems is alienating. However, that does not rebut the point that I am

making. That is, if you don’t agree with Marx on labor and its centrality to flourishing, then you

still might find it more convincing that AE’s way of treating attention is alienating.

Successfully engaging in labor depends on paying attention to something, but one can be

attentively engaged in many other activities than labor. That is, attending to things encompasses

a broader category of human activity. Unlike labor, attention is a key part of activities such as

contemplation and mindfulness, as well as simply being present with one’s friends and family. In

this way, it is not just broader but also more ecumenical. The claim that attending to things well

is key to living well doesn’t commit us to a narrow conception of what is good for people. In

fact, it is in line with many major ethical and religious traditions regarding what constitutes

flourishing, including virtuous activity, contemplation, and mindfulness. These activities all

centrally involve attending in certain ways (e.g., in a sustained manner) in order to successfully

engage in them.

While it is interesting, the Marxian point might be too narrow—laboring well in certain

ways may be one way to flourish, but it is not necessary. There are other ways to live well, and it

seems plausible that attending to things well plays a key role in this (including in laboring well).

For these reasons, attention is also more central than labor to good character development. So, it

is not as easy to dismiss the view that certain types of attention (e.g., sustained attention) are

inherently part of a flourishing life.

In general, by using attention as a main currency and central commodity, AE takes a

stand on what attention is, why it is valuable, and what its purpose is—as commensurable,

instrumentally valuable, valuable only for its quantity, and for market exchanges. But this view

might be wrong, and it certainly requires a moral defense. AE is a key example of how markets

115

are not morally neutral—that allowing a good like attention to be traded on a market, especially

as a primary currency, assumes and promotes a certain conception of the good and its value. To

challenge this conception, I provided reasons above, as well as in my other three chapters, as to

why we should instead think that attention is a key part of our autonomy, intrinsically valuable,

and central to living a good life.

Summing up, the foregoing analysis provides an important reason why AE is morally

objectionable. I argued that the attention economy, by using attention as a currency and key

commodity in widespread and central areas of life, expresses and promotes disrespectful and

degraded attitudes. As such, AE is morally corrupting. Recall that I’m providing a threshold

view—an individual attention transaction (considered in isolation) doesn’t rise to the level of

corruption.

III. A Response to Brennan and Jaworski’s Objection

In order to further defend my position about AE, I will now consider and respond to a

prominent objection to the corruption worry. In a recent article, Brennan and Jaworski (B&J)

attempt to rebut the claim that markets in certain goods express improper meanings and that this

is morally significant (B&J, 2015). Their main argument is that the meaning of markets is

contingent. What a market transaction expresses varies from culture to culture. If so, they argue,

if markets in our culture express (say) disrespect or instrumental attitudes, and we object to this

for a particular good or practice, then we should just shift the meaning to something more

appropriate. Further, if the meaning of the market cannot be shifted, then we ought to engage in

the market anyways, because the good consequences of doing so eclipse the expressive wrong

(due to its contingent meaning). So, in the case of AE, they would object that the meaning of

attention markets is contingent and that if this meaning is problematic, we ought to shift it to

116

what they call a “schmarket”. (A “schmarket” is a market that doesn’t express bad attitudes, such

as disrespect and impersonality.)

It is worth noting that there are several questions involved here. First, we can ask whether

the meaning of markets is contingent. Relatedly, one might wonder what sense of contingency is

being used. It doesn’t seem likely that those who object to markets on symbolic grounds are

thinking of the meaning of markets as necessary at (say) the same level as mathematical or

logical truths. Now, even if the meaning of markets is contingent, we can still ask whether this

meaning—within a given society—is persistent and hard to change. Finally, if the meaning is

persistent, we might very well wonder why the good consequences (such as mutually beneficial

trades) should always morally outweigh the expressive wrongs of markets in certain goods (e.g.,

sex or body parts). Let’s quickly examine how B&J respond to these questions.

B&J argue that markets have a contingent meaning, though they are not clear in what

sense it is contingent. However, in multiple places, they seem to admit that the meaning of

markets might be hard to change; and yet, they still claim (on the basis of the contingent meaning

of markets) that, when it comes whether one ought to engage in contested markets, consequences

should take priority over expressive value.

However, why should it follow that expressive meaning and value is always less

important than consequences, just because the symbol used to express it does not do so

necessarily? For example, the meaning of raising one’s middle finger is a contingent matter, but

what it expresses (in our culture) is still deeply offensive and disrespectful, regardless of the fact

that it might not have meant this. It does mean this, and this is what matters in determining

whether one should make this gesture or not. When it comes to expressive value, the connection

between contingency and moral insignificance seems tenuous at best. So, B&J’s claim that if the

117

negative meaning of a market is hard to change, then we still ought to morally downplay or

disregard it, seems implausible and to rest on an unargued commitment to consequentialism.

There are many other places where one can disagree with B&J’s argument. First, one

could argue that market transactions have some necessary meaning (e.g., commensurability).

Second, one could argue that the meaning of markets is very hard to change, especially in the

context of current markets (and AE), and that this persistent meaning matters more than B&J

seem to think. I think there is merit to the first route, and I presented some reasons for it earlier. I

will mainly argue for it indirectly here by showing how B&J’s arguments for the contingent

meaning of markets fail. Then, I will defend the second route, namely that the meaning of

markets (e.g., AE) is at least very persistent and that this carries moral weight.

Along these routes, I will now respond in detail to Brennan and Jaworski’s objection. To

begin with, B&J (2015) define a market exchange as a trade between two people for valuable

consideration (typically money for goods or services). But this is too broad of a definition of a

“market” to capture what is being critiqued about (say) attention markets. As discussed above,

the actual AE market involves a widespread social practice of many actors trading attention for

goods and services. It is not a collection of isolated exchanges, but an embedded social practice,

with a socially understood meaning and purpose. A market in this narrower sense is the subject

of prominent market critics (e.g., Sandel or Anderson), and B&J’s type of market is a broader

construct. In particular, critics like Sandel would not consider an isolated exchange of attention

for services as a market transaction or part of an attention market. To the extent that B&J are

discussing markets in one sense and critics are discussing markets in this narrower one (which

excludes, say, one-time exchanges), the two parties are not talking about the same thing.

The lack of clarity likely derives from an ambiguity or confusion in what we mean by

“market”. On the one hand, we might call an individual trade between two people a market

118

transaction if it satisfies certain local conditions (e.g. goods and services are exchanged, there is

no expectation of further interactions or obligations unless specified, etc.)—this is close to B&J’s

conception of a market. However, on the other hand, we might think that for it to be a market

transaction, it needs to take place in a larger social context, where others are engaging in the

same type of transactional interactions with the same norms (and perhaps currency). If so, it

cannot be an isolated occurrence, such as the one-time exchange of water for food by two

travelers in the desert. For something to be a market transaction, does the larger social and

institutional context matter? It seems like we sometimes use the term to describe one-off trades.

However, these trades are different than exchanges that depend (for the term “market” to be

correctly applied) upon being embedded in some larger social practice. This definition of

“market”, unlike B&J’s, leaves room for there to be exchanges that are not market exchanges.

That there is such a gap seems plausible, but let’s look at this issue a bit further.

This distinction helps to explain the problems with a key case that B&J use to support

their argument for the contingent meaning of markets. They place a fair amount of weight on the

case of the Merina people: in this culture, husbands provide their wives with money after sex.

However, this is not thought to be disrespectful, express improper distance, etc. In fact, B&J

(2015) interpret this as a case where the husband purchases sex from his wife but in a respectful

way. B&J take this as evidence that markets in some cultures don’t have these types of morally

problematic meaning, and so that markets in general don’t necessarily have them.

Yet, I ask, is this really a market transaction? For starters, it is questionable whether the

two parties (husband and wife) are engaged in an exchange as opposed to a gift. Perhaps the

money is a token of affection in the same way that (say) making breakfast or buying flowers is in

our culture. Even if this is a transactional type of exchange though, there are ways in which it

doesn’t seem to take place within a market. For something to be a market, it would seem to

119

require that there are (at least in principle) multiple buyers and sellers of a good or service. In

this case, though, no one besides the husband can purchase sexual services from the wife—there

are no third parties involved, even potentially. While this case involves a social practice, it is a

closed one, and a market plausibly needs to be relatively open (e.g., multiple actors need to be

able to buy the good in question).

At this point, B&J would likely respond that this case (of the Merina husband and wife)

is still a market transaction in the broad sense that they defined. It is still a case of two people

trading things of value with one another. They would say that markets in this sense need not have

the improper meanings that people are concerned about. But I think that this misses the point—

those who are concerned about the meaning of markets would not be convinced that this

interaction amounts to a market transaction. If “markets” are understood to exist at the level of a

social group and to require general constraints such as multiple buyers or sellers (and a

commonly used currency, etc.), the Merina people example does not support the claim that the

meaning of markets (including the meaning that they bestow to goods and services) is

contingent. Importantly, regardless of whether this holds for all markets, because the narrower,

social institutional sense of “market” clearly applies to AE, the example of the Merina people

does not provide evidence that AE markets have a contingent meaning.

Moreover, a separate argument that B&J make for their contingency thesis is flawed.

They point out how market critics, such as Sandel, are wary of marketizing social practices, and

B&J focus on his example of a wedding toast. Sandel (2012) argues that if the best man bought

his wedding toast (e.g., from a website) instead of thoughtfully writing it himself, that this would

degrade the meaning and value of the good (the wedding toast); i.e., it wouldn’t be the proper

way to express how one feels about a very close friendship. However, B&J (2015) ask us to

imagine a “Twin Earth” where wedding toasts are bought on the market and wedding cakes are

120

thoughtfully baked by the best man (instead of bought)—basically, the inverse of current social

practices in our country. In this “world”, it would be considered disrespectful for someone to buy

a cake (given the expectations surrounding the practice) but totally fine for them to buy a toast.

From this thought experiment, they infer that the meaning of markets is contingent, because it is

a contingent matter whether wedding toasts or cakes are the proper vehicle to express how much

the best man values the groom.

But this example doesn’t establish what B&J need. Even if it is a contingent matter which

social practice expresses proper respect and closeness, in both worlds the practice that does so is

the non-market one. The meaning of markets (and putting a good on the market) is actually held

constant in both scenarios; what changes is which practice is subject to the distance, disrespect,

etc. that putting a good on the market expresses and promotes. So, this does not show that the

meaning of markets is contingent; instead, at best it shows that, for some types of goods (e.g.

wedding toasts vs. cakes), it is a contingent matter what meanings are associated with them and,

thus, whether it is morally okay to place those goods on the market. In fact, this example seems

to implicitly commit B&J to the view that markets have certain meanings (such as distance and

disrespect), which is not their stated view. In fairness, it doesn’t commit them to the view that all

markets have necessary meanings, but it does seem to imply that they think actual markets have

these negative meanings and probably that they think the meaning of markets is hard to change. I

will expand on this point below.

Before moving on, it’s worth noting that, even if B&J’s case about the meaning of

wedding toasts goes through, this would still not establish that all goods are contingent in this

way. That is, this type of case might not universally generalize. Wedding toasts might have

contingent meaning—e.g., due to how they are mainly socially constructed—in a way that

121

contested goods like sex or body parts might not. In particular, there is reason to think that the

meaning and value of attention—e.g., as a core part of our identity—is not culturally relative.

I have argued that B&J fail to establish their contingency thesis. However, for the sake of

argument, I will grant to B&J that the meaning of current markets, especially those constituted

by AE, are contingent. That is, I temporarily grant that what they call “schmarkets”, which are

like markets but don’t express morally problematic meanings, could be instituted. Even if this is

so, there are still many reasons to be skeptical of B&J’s proposal for solving the corruption

worry about markets.

First, even if we could have schmarkets—if we could change meanings—the meaning of

markets is very persistent. At a much simpler level, think about the social meanings of (say)

hand gestures and think about how difficult it would be to change them. This would need to be

done over time and at the level of the social group—no individual or small subset of society

could unilaterally institute such changes. They would need to convince many others to change

their attitudes and behaviors. It is very hard to shift even simple cultural norms, let alone the

norms and meanings surrounding complex social institutions such as markets. One reason is that,

analogously to language, the meaning of these types of interactions is not up to any individual

actor. For instance, a particular hand gesture might express something offensive, even if this is

not intended by the person making it.

Consider the meaning of markets, including AE, in our society and in similar ones.

Regular market transactions in our society seem to express the meanings that market critics

disapprove of, and that B&J claim are contingent (e.g., disrespect, distance, and instrumentality).

It is important to note that part of the reason typical markets express such meanings is that they

have been set up and institutionalized to promote efficiency—any given transaction is, via

money, connected to a very large, global economy. Another feature of most markets we are

122

interested in (including AE) is that there are typically few to no barriers to entry—anyone can

trade with any other willing party. In other words, it is part of the design or structure of these

markets, and the norms surrounding how people use them, that they promote both freedom of

exchange and efficiency.

Given the design, structure, and norms of typical modern markets, it starts to look a lot

less contingent that they express certain meanings (such as impersonality). Perhaps they even do

so necessarily, given these features. But that is a stronger claim than I need here. All I need is

that in our culture, even if these markets (including AE) could express different meanings, these

meanings would be very hard to change. We certainly can’t just erase or change them by fiat.

For instance, take the corn market (or the large-scale market for attention that is AE).

Would it be easy to change the meaning of the market for corn into a schmarket (respectful,

personal, etc.) for corn? I think it would be very difficult. The actual market in corn is able to

function so efficiently by using U.S. dollars (as currency) to coordinate a wide range of

transactions between self-interested and distant trading partners. This expresses certain attitudes

such as impersonality. One way of making this market less impersonal would be to restrict who

could trade on it—perhaps only people who live near each other could make exchanges. But this

would face steep resistance, given the current norms that people should be allowed to enter and

exit typical markets (e.g., for corn) freely, and given that it would generate inefficiencies (since

willing parties who were geographically separated would not be able to transact). At both the

personal and political level, it seems likely that people would not accept such restrictions (and

their consequences).

And even if such a restriction to entry was somehow successfully instituted, money

would still be the currency. U.S. currency can be traded for many other goods that are sold on

markets open to an unrestricted number of participants. In this case, it would make it hard to

123

prevent these restricted corn exchanges from expressing estrangement, because money would

still connect such exchanges to a very large and impersonal group of people.

Well, someone might say, perhaps the next step would be to try to eliminate the use of

U.S. dollars as currency and institute (say) a barter or IOU system (or even a special currency for

corn markets). But all of these proposals would further limit the number of participants and

generate even larger inefficiencies. Beyond a very limited group of trading partners, it would be

harder to coordinate exchanges, to trust that you would be repaid, to access the larger economy,

etc. Efficiency and freedom considerations are a large part of why we have one general currency,

and not a specialized currency or a barter system for each individual market. If restricting access

to the corn market would be difficult for people to accept, attempting to disallow U.S. currency

would be even harder to institute. Moreover, what would it look like if this could be instituted?

Because of the features it would need (strong barriers to entry, a different means of

exchange, etc.) it is hard to imagine what a corn schmarket (respectful, personal, etc.) would look

like in modern society, at any type of large scale. Perhaps what this shows is that schmarkets are

only feasible (or even possible) in small societies, or in very niche, isolated corners of modern

ones; and not at the level of national or international economies. So, the corn market example

provides further reason to think that the meanings attached to markets in our society, such as

estrangement, are, at the very least, not malleable and highly persistent.

Here's another, related, reason to be skeptical of B&J’s proposal—namely, they wouldn’t

want to change AE markets into schmarkets anyways, because of their moral commitments to

consequentialism. Recall their claim that, if we can’t change the meaning, then we still ought to

participate in markets that have an improper meaning. This move depends on the assumption that

consequentialist considerations take moral precedence over (“mere”) expressive ones.

124

Suppose that we could institute schmarkets in corn (or any other good) that were

respectful, personal etc. If we were to do so, then there would be substantial dead-weight loss

(unrealized gains from trades), which is at odds with B&J’s consequentialist, efficiency-based

commitments. Why would there be dead-weight loss? Well, for instance, there would likely be

cases where some people would want to trade their corn for valuable consideration (money or

services), but where the norms of exchange (based upon respect, etc.) would deter buyers and

sellers from making such trades. They would refrain because of moral considerations like respect

and others we’ve discussed. Imagine ice vendors after a hurricane refusing to raise their price out

of respect and personal connections to their customers. They wouldn’t want to exploit or price

out their close connections. But this treatment and attitude would result in a shortage—there

would be more quantity demanded at this price than quantity supplied. Some people who would

be willing to buy at a higher price would be unable to do so. But this is dead-weight loss—both

parties lose out on the potential gains (in terms of preference satisfaction) from making a trade.

Additionally, it is likely that there would need to be some amount of government or

social regulation to help change the meaning of markets into “schmarkets”. That is, some amount

of external enforcement (e.g., regarding price increases or means of exchange) is probably

needed to help override the prevailing norms of current markets. If so, this increased limitation

on freedom of exchange would surely be undesirable to many market defenders, including B&J.

Thus, both because of its difficulty and its undesirability (for its very proponents), the

whole proposed solution of shifting the meaning of markets to schmarkets is a red herring. It is

not a realistic or feasible solution. And B&J should reject it anyways because of their own moral

commitments, including those expressed in their very same paper—i.e., that consequences take

complete priority over expressive value. Given these commitments, B&J should balk at

125

instituting schmarkets in particular goods (even if it were possible), because schmarkets limit

one’s freedom to exchange and they fail to yield efficient outcomes.

As a final consideration, the difficulty of shifting the meaning—due to persistent features

of actual markets—is a concern that applies even to those who do not share B&J’s moral

commitments. That is, changing the meaning of markets would be a difficult endeavor, even for

those very worried about improper expressive value and its close connection to people’s

character and attitudes. Once we have a market in something, at least at a large scale and in our

present culture, it is very hard to escape the problems raised by the corruption objection. In

particular, the degraded attitudes that markets in attention express and promote are subject to

these problems. We certainly cannot switch the meaning of markets—for attention or in

general—off and on like a light switch.

My rebuttal of B&J applies to many typical modern markets, not just AE. Importantly,

though, AE is included. I provided strong reason to think that—along with other large-scale

markets—the meaning of AE’s market in attention is (at a minimum) very persistent, and also

that B&J’s arguments for its contingency are flawed. Thus, my argument that AE expresses and

promotes degraded attitudes about attention emerges unscathed from B&J’s objection.

IV. Conclusion

In this paper, I argued that AE expresses and promotes degraded ways of thinking about

attention—namely, that it is commensurable with other goods, instrumentally valuable, and

primarily to be used as a means of exchange. I argued that these ways of thinking about attention

are morally degraded for multiple reasons—including that they disrespect the intrinsic dignity

and value of a person, they incorrectly treat attention as only valuable for its quantity, and they

are at odds with both autonomy and living well, since they downplay and downgrade the value

126

and role of sustained attention. In the process, I also argued that the scope of AE is large and

pervasive, which enhances the significance of the moral harms I analyze. Then, in section III, I

defended my view from a prominent objection raised by B&J. Their argument for the contingent

meaning of markets is flawed, and the persistent (if not necessary) meaning of typical modern

markets, including AE, carries significant moral weight. Despite their objection, the argument I

make for the expressive, attitudinal, and character-based moral wrongs of AE still holds.

127

REFERENCES

Anderson, Elizabeth. 1993. Value in Ethics and Economics. Harvard University Press.

Aristotle. Nicomachean Ethics in Readings in Ancient Greek Philosophy 5th Edition. ed. Cohen,

Curd, and Reeve, C.D.C. Hackett, 2016.

Brennan, Jason and Peter Jaworski. 2015. “Markets without Symbolic Limits.” Ethics 125: 1053-

1077.

Castro, Clinton and Adam Pham. 2020. “Is the Attention Economy Noxious?” Philosophers’

Imprint, 20, no. 17: 1-13.

Plato. Republic. Translated by C.D.C. Reeve. Hackett, 2004.

Sandel, Michael. 2012. What Money Can’t Buy: The Moral Limits of Markets. Farrar, Straus, and

Giroux.

Wolff, Jonathan. 2002. Why Read Marx Today? Oxford University Press.