All Benefits Should Be Portable Benefits
I sent some comments on a prepublication draft of this piece to Shelby in email, which I wanted to share more broadly here. I wrote:
“A couple of thoughts, which are not suggested emendations, just additional perspectives:
One way to solve the tradeoff between employment security and flexibility is to add portable benefits to an on-demand independent contractor model (or to create some kind of hybrid third classification.) And it’s good that so much of the focus has gone there. There are many obvious benefits to on-demand labor for both individuals and companies. But companies have abused the flexibility of this approach primarily to reduce costs rather than to enhance the benefits for employees. Required portable benefits, with no distinction between part time and full time work, would remove the elements of cost savings that come from reduced benefits, and focus on the cost savings that come from more accurately matching labor supply to demand. In some ways, erasing the distinction between full time and part time is a precursor to portable benefits, and without it, they won’t work, since without it, employers would be incented to limit working hours in order to get under a benefits cap.
Another approach would be to add the flexibility of an algorithmic labor market to traditional employment. Imagine, for example, if instead of the tightly scheduled shift work at retail and fast food outlets that I outlined in Workers in a World of Continuous Partial Employment as the existing labor market alternative to on-demand companies, there were an Uber or Lyft equivalent that would allow McDonalds or Burger King to post available shifts to their committed employees, and let them choose which ones to take. You could imagine an interesting realignment of the low wage job market if it were a real marketplace, where companies, for example, had to offer higher wages to get people to take undesirable shifts rather than just assigning them as a condition of employment. It might get even more interesting if this were a marketplace that gave workers visibility into the shifts available from all employers offering similar jobs. You would still need the protection of portable benefits to make sure that there were no incentives to limit hours.
What this would ideally require was an open standard for scheduling information, a requirement to make it accessible to a marketplace, and of course, the labor marketplace itself, so that workers could see shifts at say, both Burger King and McDonalds and choose whichever paid the most or had the best time slots. If I were one of the big workplace scheduling companies, I’d be working on this end of the problem.
There are some steps in this direction.
Pared is a company started by restauranteur Will Pacio to address the problem faced by independent restaurants who are trying to keep fully staffed. It’s an on-demand network for jobs like sous-chefs, dishwashers, and other “back of house” jobs in the hospitality industry.
Now imagine that this marketplace were used as a focus for creating some market power for workers, so that it wasn’t big companies facing off against powerless individuals but against a modern organizing force. I rather like what both OurWalmart and CoWorker.Org have done in this regard. They aren’t doing traditional organizing, but they are using the internet to increase worker voice.
There are other industries that could also benefit from this approach, such as home health care. Honor is a company that provides on-demand home care for the elderly and others in need of one-on-one attention. They started out with 1099 and switched to W2, but still give caregivers affordances for flexible schedules. But it may be hard for them to scale unless they adopt a marketplace approach.”
Shelby replied:
“My personal perspective is that I’d rather see a push to a single employment classification (there being no difference between IC and employee) than moving towards a third classification. At a minimum, I believe that certain basics, like workers comp or disability, must be provided to all workers, and then we should open the door for companies to provide additional benefits, like health, or retirement. A framework similar to the Black Car Fund in NY could provide this.
Perhaps naively, I’d love to see a world where the on-demand companies will start to compete for high-quality supply thru benefits, instead of the McDonalds example where they come up with every trick in the book to avoid benefits. I do agree that if we create a viable system of benefits, it could have a great on w2 employees by creating a new linear paradigm instead of the binary part-time/full-time model, which creates strange incentives. I’m also very optimistic that, if done right, a portable benefits system could be a very strong mechanism to provide worker power. Solve a problem for workers (affordable benefits), and you have organized them. This is similar to the Ghent System, which we provide in the paper, which is the Nordic Union system where workers voluntarily join unions to access unemployment insurance, which has led to the highest unionization rates in the world.
Hopefully we’ll start to see some concrete proposals on portable benefits begin to emerge — exciting to begin thinking about where this could evolve — as you’ve begun to help us visualize.”
I very much agree that an essential step towards portable benefits is to blow up the whole W2 vs 1099 classification, or restrict 1099 to true arms length compensation arrangements (for example, an advance on royalties given to a book author, or dividends on stock ownership). Otherwise, there are indeed very strange incentives in the part-time/full-time distinction. As I discussed last year in the piece that I link to above, low-wage W2 employers are incented to keep employees below 30 hours to avoid additional benefits, which is bad both for workers and for employers.
The Aspen paper that Shelby links to here is a great roundup of practical work that has been done on portable benefits, and is required reading for anyone interested in the subject.
