Reflections on the sharing economy
March 22, 2019
If you are getting your work through a platform like Postmates, Uber, or Lyft, or hosting on Airbnb, chances are there’s a college professor or PhD student who wants to talk to you very badly. That’s the conclusion I reached this week after attending the SHARE Conference put on by a team of researchers at Boston’s Northeastern University. Share is an acronym for Sharing Economy-Humans, Automation, Resilience.
One of the keynote speakers, Daniel Castro, noted that according to Pew Research 73% of Americans are unfamiliar with the term “sharing economy,” and there is debate about what the term means even among those who study the topic.
Castro also acknowledged that the sharing economy is not new, citing libraries and gyms as some of the earliest examples of sharing goods. His preferred definition is the “use of digital platforms to match fair capacity and spare demand.”
Here’s a round-up of the topics I found especially interesting:
Free Riders Beat Out Full Timers
A new study from a team of Boston College researchers argues that platforms like Uber, Lyft, and Postmates have failed to create a viable model for full-time self-employment for younger workers. In a study funded by the Macarthur Foundation, sociologist Juliet Schor and her team explored seven platforms over a period of seven years, conducting in-depth interviews with 111 workers aged 18 to 34.
“In the platform model, the employer cedes control over hours and some aspects of the labor process and greatly reduces barriers to entry,” Schor said. “We need to appreciate how radical this is. The result is that there is far more diversity in the labor force than conventional employment, and we think that makes a big difference.”
We need to appreciate how radical this is. The result is that there is far more diversity in the labor force than conventional employment, and we think that makes a big difference
Schor’s research found that relationship to the platform really influences outcomes. Those using it to supplement their income have better outcomes than those who are entirely dependent on it for their livelihood. “Dependent earners are generally not achieving adequate incomes, security or benefits,” she said.
While other academics argue that outcomes of the workers are heavily influenced by policy decisions governing employee classification and provision of benefits, or ceding control to machines for algorithmic-driven assignment of work, Schor’s team found that the low amount of control that platforms exert over its users creates a diversity of ways that work can be arranged or conducted.
“The platform labor model is a significant departure from previous labor control regimes. What our findings suggest is that there are ways the platforms are very significantly giving up on controlling things, and that has big implications,” she said.
In the study, 42% of participants were supplemental earners, with the highest proportion from AirBnB and TaskRabbit. She said the preponderance of supplemental earners on platforms creates a “free-riding” effect, where their primary W-2 employers are subsidizing the platform.
“Their (the platform’s) success in attracting and retaining labor is in large part due to the contributions of W-2 employers to the incomes, benefits, security and well-being of many platform earners,” she argued.
The Algorithms are Messing with You
While gig economy companies run ad campaigns touting the freedom and flexibility of self-employed life, the reality, according to Lindsey D. Cameron of the University of Michigan’s Ross School of Business, may be a little different when the work process itself is controlled by machines and algorithms.
In her quest to understand the way algorithms and the humans who work with them interact, Cameron went as far as to work 150 hours as a driver and conduct almost 200 interviews with both ride share passengers and drivers. According to Cameron, eighty percent of jobs of the future are expected to involve algorithmic work.
Platforms rely on algorithms to structure the work process. “You may give 10 rides but have over 100 interactions with an algorithm,” Cameron said, noting that it’s possible to do work for a platform without ever meeting a single employee of the company.
Her conclusion? Workers have what she called “contingent autonomy” versus the algorithms, meaning that some activities, like deciding when to drive, are at the discretion of the driver. Others, like work matching, are more tightly controlled by the service. “Autonomy is no longer static within a role,” she said.
Autonomy is no longer static within a role
Sharing in Name Only
Will Attwood-Charles of Boston College studied couriers from the platforms Postmates and Favor, conducting in depth interviews with 25 of them to understand whether gig workers can develop a shared culture or identity.
Attwood-Charles noticed three primary ways that couriers defined themselves relative to the job:
- Aspirational identity and role conflict, which most often involved someone working part time on the platform while they trained to do something else or worked at getting another profession off the ground.
- De-identification and entrepreneurialism, which involves seeing oneself as an entrepreneur who is motivated by the idea of starting a business with the platform serving as a means to an end.
- Oppositional identity formation, which is based on shared grievances, usually related to how the platform is taking advantage.
Taken together, these studies gave me a lot to consider regarding our future relationship to work and each other. I’m comforted by the fact that social scientists are actively exploring the platform phenomenon while it’s still in its infancy. There is still time to do something before the machines take over.