Wednesday, April 08, 2020

Reading :: Heteromation

Heteromation, and Other Stories of Computing and Capitalism
By Hamid R. Ekbia and Bonnie A. Nardi


This book's central argument is that "digital technologies are responses to the predicaments and opportunities of capitalism" (p.7). "While others debate whether digital technologies are mere tools or drivers of social change, we regard them largely as a response to the dynamics of socioeconomic change" (p.8).

Based on Marx, they argue that labor is "a uniquely human capacity, which accounts for the production of economic value in capitalism" (p.12). In this context, they introduce the concept of heteromation, which is "the extraction of economic value from low-cost or free labor in computer-mediated networks" (p.xv). They ask:
  1. What are the social and technological processes through which economic value is extracted from digitally mediated labor?
  2. What is the nature of the value created in this process to allow capital to continue its necessary expansion?
  3. How are people incited to participate in this process? (p.3).  
 They argue that heteromation is "a capitalist innovation to deal with the tension that derives from [the] paradox" that Marx described, in which the rate of profit tends to fall due to "technologically driven rising productivity" (p.16). Heteromation represents "a new law of capitalist accumulation" (p.17) in which economic value is extracted "from uncompensated or low-wage labor, inciting participation through an intricate set of mechanisms comprised of social and emotional rewards, monetary compensation, and coercion" (pp.24-25). It extracts this labor via inclusion, engagement, and invisibility (p.39).

The authors deploy a variety of examples to make their argument: Mechanical Turk, social media use, kiosks, self-serve checkout at grocery stores. However, I found these examples to be too dispersed to cleanly make the argument the authors want to make.

For instance, take self-serve checkout—which (as the name suggests) disaggregates a component of service from the purchase of goods. The authors say that "a close look reveals that a good part of the labor is, in fact, done by another group of human beings—the consumer, the end user, an intermediary such as a family member, or in some cases, a new kind of casual laborer. This labor costs capital little or nothing" (p.108). Although automated grocery checkout is said to replace humans, the authors argue that "machines are not fully replacing humans, but rather reconfiguring the labor into the heteromated labor of end users who do the cognitive work" (p.110). The authors conclude that this work is "oxymoronically" called "self-service" when it is "no service at all" and instead hides the labor that the consumer ends up taking over (p.111).

But this argument seems to assume that service labor is an intrinsic part of what people are paying for, and thus they are being tricked into performing that labor themselves, while the company keeps the profits of their labor by charging for service that it does not produce. It's possible to view things differently: heteromation, in some cases, allows the provider to partially disaggregate service from the cost of goods. That is, before heteromation, the grocery store had to integrate the service of checkout into the cost of goods because otherwise people would walk out without paying, or pay the wrong price, or otherwise elide the transaction. With heteromation, some grocery shoppers are willing to perform that labor themselves, lowering labor costs in a business with notoriously thin profit margins. The service is no longer part of the goods. Consumers are evidently okay with this tradeoff—how many of us, given the chance, would choose a full-service gas station over a self-serve after comparing the price per gallon? For that matter, how many of us eat every meal at a restaurant (where the cost of labor and service is included in the price of the meal) rather than buying our own groceries and performing the labor of cooking at home?

The authors' argument also doesn't really engage with the relational aspect of value propositions (and I guess I would argue that Marx, who was specifically focused on commodities, doesn't either). For instance, the authors wrote this argument before COVID-19, and I'm writing this review afterwards. COVID-19 is a nice example of how value might change based on circumstances: I'm more likely to use, and see value in, a kiosk vs. a cashier now because the kiosk allows me to actively avoid germy people (I still have to touch the same germy keypad). What was an imposition in January is now, in April, an additional service, and I'm likely to seek out grocery stores that can provide this service. That is, heteromation is preferable for me and I might actually pay more for it because it provides the service of keeping me safer (just as people sometimes pay additional service fees to use ATMs).

The picture is more complicated, of course. Grocery stores still employ checkers, just not as many. They don't (yet) have a differential surcharge for human service. And automated systems do still have tenders, but the ratio of tender to checkout machine is more like 1:4 or 1:8. Finally, the provider may still keep charging for service that it is no longer providing—but to make that case, one would have to carefully disaggregate and examine these costs, and the authors don't do that.

In any case, this case—in which service is steadily disaggregated from cost of goods—seems rather different from the case of social media, in which people create value for the social media company simply by using its product. Or Mechanical Turk, in which people perform casual cognitive tasks.  Ultimately, I wanted the authors to make a stronger and more nuanced argument that these different cases did fit together, that they represented the same overarching trend, and that they all counted as labor.

As you can tell, I have a lot of thoughts about this book. And for that reason, I recommend it—although I was not convinced by the argument, this argument gave me plenty to think about. The authors draw broadly on various literatures to define and describe heteromation, a concept that can underpin other investigations into distributed labor.

No comments: