For two years I have worked with a research team studying the lives of crowdworkers, people doing online jobs that can entail everything from flagging tweets to transcribing a doctor’s visit. We surveyed thousands of people doing this new form of digital piecework, both in the United States and India, interviewing hundreds of workers, in their homes and at local cafés, as they did their work.
Mary L. Gray is a senior researcher at Microsoft Research, associate professor of the Media School at Indiana University, and a fellow at Harvard University’s Berkman Center for Internet and Society. She is writing a book, with computer scientist Siddharth Suri, on platform economies, digital labor, and the future of work.
These workers scoured websites and LinkedIn profiles for the best company contacts so that salespeople could more quickly reach the person with the power to say yes. They looked at thousands of pictures, assigning them labels like “family” or “amusement parks” or “camelback couches,” to correct and improve the algorithms used to show us relevant images when we use a search engine. They typed in information from business cards and handwritten notes so that it could be neatly stored and readily retrieved electronically.
We met people like a college-educated, stay-at-home mom staving off boredom while her children are at school, and a first-generation college student, working 50 hours a week to save for his wedding and help pay for his younger brother’s degree in finance. And a host of other people spending two to 70 hours a week doing work that makes the Internet seem magically automated yet remains largely unappreciated by many of us.
Crowdwork represents a small but growing microclimate in the ecosystem known as platform economies. These are business activities burgeoning through the ties that bind the Internet, smartphone apps, and social networks. The boss here is not a mid-level manager but an application programming interface or API deployed on a platform. The platform—owned and operated by companies like Amazon, Upwork (formerly oDesk), and LeadGenius—works in concert with an API to generate and verify workers’ accounts, handle the workflow of millions of online job postings, and route payments to workers once they complete their tasks and submit them for approval from an invisible, ethereal “employer.”
In many ways, the assumption that workers no longer need a supportive or collaborative work environment and can act as self-directed actors is a fair one. Plenty of workers figure out how to find a good gig, develop routines for getting work done quickly, even find water-cooler chatter on worker-centered forums. A significant percentage of crowdworkers string together 30 to 50 hours of work, and rely on networks of peer support to maintain this level of productivity. Workers share information about how to sign up for platforms, what jobs to consider, employers to avoid, even how to do certain tasks when the task instructions leave out key details. Indeed, my time with workers shows that the API silently shifts the burden of finding, training, and retaining talent to workers’ shoulders.
Helping these workers connect—fostering rather than ignoring or stifling their collaboration—is the right thing to do, and the companies operating and profiting from the platforms that generate employment are in the best position to do the greatest good for workers. Providing online discussion forums and offline meet-ups, and rewarding more experienced workers for showing newbies the ropes, likely generates higher-quality work. We also found that the high attrition of labor in these systems comes in part from individuals who never find their way to workers’ forums or the personal networks of other individuals working on these platforms.
Crowdworkers today bear an unfair share of the hidden costs of this new economic engine. We must re-imagine a global social safety net that provides workers with a basic income and health benefits detached from the institutional digs and hours they work. We must better reward workers willing to take on a task at a moment’s notice. We should look to profit-sharing, higher corporate taxes, and higher capital gains taxes on investments in these companies to help subsidize this on-demand workforce willing and able to contribute any time, day or night, to the growth of this economy.
We also need a third-party registry that allows crowdworkers to build their work resumes and accrue reputations, independent of the platform, that they can then take with them, no matter where they pick up their next gig. And, lastly, consumers and workers have a right to know when a platform “games’’ its search results. In other words, we need to know that when some workers are more prominently displayed it’s not necessarily because they did a better job but because the platforms sold their top search-result placements to the highest bidder.
While we must develop robust mechanisms that prevent individuals from scamming platforms in the on-demand economy, we must with equal vigilance penalize employers for misclassifying, delaying, or failing to pay workers, one of the greatest challenges facing those making a living at freelancing today.
Supporting the many people who may never enjoy the security of a 40-hour workweek will be one of the most important conversations we have about the on-demand sharing economy.
Editor's Note: A version of this post originally ran on Bloomberg View.
For the Future of Work, a special project from the Center for Advanced Study in the Behavioral Sciences at Stanford University, business and labor leaders, social scientists, technology visionaries, activists, and journalists weigh in on the most consequential changes in the workplace, and what anxieties and possibilities they might produce.