Because: capitalism! is not an argument by itself

Students in my class will often blame "capitalism"--by which they often mean "profit-seeking"--for particular outcomes they don't like. But there is something more fundamental: interest-seeking.

In my “Introduction to Science, Technology, and Society” course, I spend about five weeks teaching students about how social factors shape technological systems. These can range from designer contexts, i.e., the constraints that designers face in terms of financing, available technologies for manufacturing and such; to questions of maintenance and standards; to adoption, or what people (“users”) actually do with technologies when they get into their hands.

In the final week of the course, I talk about the question of “user representation” and its increased salience in many tech workplaces. What does this mean? Technology companies now want to build products that are “user-centered,” that take the voice of the user into consideration.¹ The goal is to make products that work better.

But what does it even mean to incorporate the voice of the user? As with the “voice of the people,” the problem is that users do not speak in one voice. Different users have different concerns and want different things. And there is just no question that technology designers can go out and speak to ALL users. So what often happens is that they end up sampling their users; and they end up using a variety of techniques. Sometimes, the technique is A/B testing, running an experimental test in which two groups of users, often unknowingly, are given two different features and then measured in terms of how they use them (with the goal of ultimately picking the feature that works “best”). At other times, the technique can be interviewing or surveying users about what they want and need. And at still other times, users themselves—or to be more precise, a faction of users—will organize and advocate for something they clearly want (on which more below).

All of which leads to the biggest problem: what does it mean to channel the voice of the user? And when you do, how do you sample and what methods do you use to “listen” to users?

This was the situation that Instagram faced a couple of years ago. In The Washington Post, the journalist Taylor Lorenz wrote about Instagram’s crisis of identity. Instagram had begun as a way for people to share photographs with their friends. But as it grew and as it succeeded, it had become a sort of platform for people to consume photographs from actors, musicians, sports people, and many others; a new category of people became “influencers,” people who made their living through Instagram by sharing content. And then Instagram faced competition from TikTok which made no bones about the fact that it was not a platform to share with friends; but instead a mass entertainment channel of influencer-generated videos, rather than photos. This meant that Instagram had to respond in kind: it had to change its recommendation algorithms so that people were given content from their non-friends that they might be most likely to engage with and second, to focus on videos rather than images.

illumitati

A post shared by @illumitati

Fair enough. But some prominent users of Instagram objected. Surprisingly, these were Instagram content creators with large followings. They wanted Instagram to be like it was before: more personal, with more emphasis on photos and followings rather than recommended content that came outside the content you were choosing to follow. These high-profile creators started Instagram campaigns and petitions. Their goal was ably summarized by their slogan (see above): “Make Instagram Instagram again (stop trying to be TikTok, I just want to see cute photos of my friends.”

mosseri

A post shared by @mosseri

These rumblings became a cause of concern to the Instagram leadership—you don’t want your lucrative creators to be mad at you—and Adam Mosseri, the chief of Instagram, responded with his own video message (see above). In the video message, he attempts to pacify the disgruntled creators; some of the things they are complaining about are just tests, he says. But he also pushed back, arguing that Instagram’s data showed its viewers responding with their feet, consuming more videos that photos. Instagram viewers liked videos, Mosseri said, and they spent more time watching videos than they had with photos.

All of this represents a conundrum. Which Instagram users should be listened to? Lorenz’s piece ends with a great quote:

“There’s a war between people who want Instagram to be more like Snapchat and people who want it to be more TikTok,” Woodbury said. “Right now the former group is larger and louder.”

So I asked the students to work through an example in small groups.

Imagine you are people in Instagram product teams. You have different ways of “listening” to your “users.” Rank the three in the order you will prioritize and say why. 

1. A/B tests and data analysis

2. Change.org petitions

3. Campaigns started by Instagram posters with large followings

By asking students to pretend to be “Instagram product teams,” I was trying to structure their inquiry. My goal was to show them that even for Instagram, there were many different choices to make but they all depended on making an active decision about prioritizing which users to listen to and how to listen to them. There was no one homogenous user. There were different users who wanted different things. The “user” to be listened to was as much a product of designer contexts and methods as he was a real person.

But in the debrief, one student raised his hand.

He said that he thought we shouldn’t listen to any of these users. Users didn’t necessarily know what they want, for e.g., people often ate foods that could end up harming their health in the long-term. Similarly, Instagram was not a useful outlet of people’s time.²

His point was: why were we even discussing these different ways of prioritizing users when Instagram use was harmful and immoral? Instagram was a for-profit enterprise that made money when people posted and watched photos and videos. But this consumption of photos and videos was bad and Instagram made profits by encouraging more consumption.

Another student took up this line of questioning and connected it back to the question of the “voice of the user.” She argued that, in a sense, the problem of listening to the voice of the user—the tensions in this undertaking—were dwarfed by the fact that Instagram made its profits by getting people to watch videos. Or to put it differently, the very strong implication was that once the profit-seeking was eliminated, the problem of listening to the voice of the user would disappear, or at the very least, not be so interesting anymore.

It is an interesting argument. But is it correct?

I think the students here were engaging in the “Because: capitalism!” fallacy. I take this phrase from a post written by Henry Farrell on the Crooked Timber blog, part of a seminar on David Graeber’s Debt that went disastrously awry (although with the passage of time, there is much there that is interesting).

Farrell wrote a post that was quite critical of Graeber’s theory for why the US dollar was the hegemonic currency: Graeber believed that this came from the US’s military power; America threatened to invade countries unless they bought T-bills (and the Iraq war was an example of that). Farrell replied that while the American empire had a variety of mechanisms that sustained it, Pax Americana did not exist simply because the US threatened to invade countries that did not pay tribute; there were other causes of state action. Graeber did not like this response; he thought Farrell was too timid to acknowledge US imperialism (among other things).

In the follow-up to the follow-up (I told you there was drama), Farrell used the phrase “Because: imperialism!” to characterize Graeber’s response to his response . Graeber, said Farrell, assumed that empire was the reason for most actions taken by the US and other states, and that there was just ONE mechanism of empire, the threat of full US military action. So Graber ended up often taking quite autonomous actions by state and economic actors as responses to US imperialism.

Hence “Because: imperialism!” In Graeber’s worldview, if US imperialism (through the projection of US military force) didn’t exist, most problems of the world economy would not exist either.³

David Graeber's theory of technology

The two students were making a similar “Because: capitalism” point. Everything they found problematic about social media—that it was a time sink, that it lacked gatekeeping and became the venue of the worst kind of rumor-mongering and falsehoods, that it had become a fountain of inauthenticity—was explained by the fact that social media was a profit-seeking enterprise.

But, in the specific question I had posed to the class, they were also arguing that the problem of understanding what “the user” wanted would also be much simpler in a world without profit-seeking. It was profit-seeking, in other words, that made the “voice of the user” so problematic and so difficult to get right. Once the profit-seeking was out of the way, it would be much simpler.

But would it really? Let’s take an example where profit-making was not a consideration.

A network to facilitate central planning. Source: Aeon.

In “How Not to Network a Nation,” the historian Ben Peters tells the story of the failed Soviet attempt at creating a computer network. Now, the Soviets had constructed a society that was built around state ownership of all the means of production; it was therefore characterized by a complete absence of what one might call profit-making. The Soviets had in mind a network that would connect all their factories and government departments together. The departments would know in real time all the metrics about production; they could then carry out their centralized planning much more precisely and set daily and monthly targets with more accuracy.

And what happened? Was the dream of more accurate central planning achieved? Peters tells us that the network itself was never built to fruition. Why not? It died as the different departments of the Soviet government waged a war about who would build it and manage it and what it would be used for. As Peters puts it, it “floundered due to unregulated competition and institutional infighting among Soviet administrators.”

The point here is that the different government agencies and factions involved in building a computer network had different interests. Those interests, in some sense, dictated the approach they took towards its structure and its workings. And because these interests were irreconcilable—the squabbling never really ended—the project itself died.

Peters has a great paragraph on his book describing what organizational life was like in the profit-less Soviet system:

The Soviet system […] sought to produce one solitary good above all else—the political good of a life apart from the capitalist experience. In a narrow sense, it succeeded: the marketplace of Soviet economic interactions became foremost a negotiation of political power rather than price. Its bureaucrats bowed to unintended incentives to exploit the rampant organizational dissonance that they oversaw, its technocrats lived by their social wits, and the system squeaked by on the capricious politics of planning run amok.

In other words, the politics that would ordinarily have taken place around the price of a commodity were replaced by the “capricious politics of planning run amok,” as bureaucrats “lived by their social wits” and attempted to exploit the “organizational dissonance” of the labyrinthine Soviet government system.

The point then is that the profit motive is only one of the several interests that operate around technology development. It’s an important one in a economically liberal capitalist society. Edison’s search for the best filament to build an electrical lighting system was guided by a key insight: that the price of his electrical lighting system should be equivalent to the cost of gas. The gas refrigerator failed because its investors were too timid compared to those who invested in its competitor and eventual winner, the electric refrigerator.

But the profit motive is not the only interest at play. When some doctors tried to create an electronic text messaging system to replace paging, it did not take off because doctors using the system had different interests based on whether they manned the ER or whether they were consultants.

Instagram content creators, all profit-seekers, have different interests as well: some want to go back to a time when Instagram was about photos and friends (which benefits them) while others want an Instagram oriented around videos and recommendation algorithms (which benefits them).

It is tempting to think that all conflict will disappear if the profit motive is taken away. “Because: capitalism!” thinking often carries a whiff of old-fashioned moralism. If only we didn’t rush after money, the world would be a much easier place. But that’s not really true. (Which is not to say that there is no place for institutional guard rails that keep the profit motive in its place.) Rather than looking for an easy abstraction (“capitalism,” “imperialism,” “colonialism”) to explain unfavorable outcomes, analysts must go deep into the conflict of interests that lead to particular consequences.

1. This is a movement that started in the software industry but has also reached other industries that make consumer products. At least a little bit of this movement that originated in Scandinavia, called “participatory design,” was motivated by leftist considerations of listening to the worker. But it is a movement that was also rooted in market factors: once the PC became a mass product and once software moved into the network, it became more and more difficult to think about how it would be used than when software was just a CD in a box.

2. I like ribbing students. I asked him if what he meant was that it was better to make enlightened decisions for users rather than trying to listening to them, however fraught that might be. He grimaced and said that sounded like fascism. There was some laughter.

3. This actually fits in well with Graeber’s theory of technology in which technology development is merely the consequence of decisions by the “ruling classes.” Graeber gives certain powerful actors, like the US government, a lot of agency.