Employee vs Contractor: The Role of Data and Getting Past the Distinction

I have written in the past about how having richer information channels is letting us overcome various historic dichotomies. For instance we had to distinguish between private and public companies because disclosing information was costly. We relied entirely on the signaling of degrees, but now in many fields people can easily show their work directly. The worker-contractor distinction is similarly the result of not having enough realtime information sharing.

There was a time when the overhead of tracking every work hour was high and at that time it made sense to treat employees who normally work for one company for a long period of time differently from contractors who move around a lot. Today the cost of metering work is virtually zero. The same goes for paying taxes frequently as opposed to once a year or possibly quarterly.

Since there is no clear distinction between workers and contractors, we had come up with a set of criteria for drawing the dividing line. Among those criteria was a notion of who controls which tools are used to carry out the work and who pays for these tools. According to the law you are more like an employee if the company requires specific tools which it provides and more like a contractor if you are providing your own tools.

In a world of digital technology of course the tool itself becomes largely irrelevant. It doesn’t matter if you are using an iPhone or an Android phone. Or if you are on a MacBook or a PC Laptop. All computers can do the same thing. Instead of the computer or smartphone what really matters is whether or not you own the code that is executed and whether or not you control the data.

So if we are going to maintain the dividing line for now, then a key consideration should be who controls the code and owns the data. If the company owns the data of the service and you can only provide the service if you are executing the company’s code, then this should be considered a factor that points towards employee status. Conversely if companies provide an API so that the work can be performed using one’s own software and one has access to one’s own data then that should point in the direction of contractor status.

In the long run though our goal should be to do away with this artificial distinction between employees and contractors also. If you had a Basic Income then one wouldn’t need payroll taxes at all and income tax could be collected as people earn additional income (any dollars earned above the basic income should be taxed). There would then be no need for a complicated yearly tax filing process and also no working capital problem for government.

All of this kind of thinking is premised on the principle of “don’t automate, obliterate” – too much of what is currently being debated in the policy realm is about automating existing processes and further enshrining categories that made sense historically but no longer do. It is time to figure out what government and society can and should look like now that we have digital technologies.