Guy Davidov is the Elias Lieberman Professor of Labour Law at the Hebrew University of Jerusalem, and currently a visiting scholar at Berkeley Law.

This post is the first in a three-part series.

The status of Uber drivers – and others who work for on-demand platforms – has been the subject of a heated debate (and of course an elaborate and highly useful discussion in this blog).  Are they independent contractors, as argued by the firms operating the platforms, or employees?  The recent settlement in one of the suits, in which the plaintiffs agreed to remain independent contractors in return for other concessions, does not settle this crucial issue.  Other suits claiming misclassification remain active.  And crucially, the normative questions are still being debated. In the current post and two following ones I focus on the normative question of what the law should be in this regard (rather than the current state of the law).  Before directly addressing the question of how such workers should be classified, it would be useful to raise a few preliminary questions.

First, does it make sense to retain the employee/independent contractor distinction?

Problems of deliberate misclassification, as well as hard cases in “gray areas” between the two poles, are hardly new.  Such difficulties have always been an inseparable part of the law in this area, and they have exacerbated over the last three decades together with the proliferation of atypical, non-traditional (and usually precarious) work arrangements.  The “on demand” economy may be revolutionary in some respects, but in the current context the challenges it creates are just more of the same.

Given the ongoing difficulties, several scholars have argued in recent years (before Uber came to existence) that the distinction should be abolished and the protections afforded by labor and employment law extended to independent contractors (and possibly others) as well.  However, while it may well be justified to redraw the lines, some line-drawing is unavoidable when setting the scope of labor and employment protections (as is often the case in law).  Moreover, although by dramatically extending coverage (for example to all those who perform work for others) we can minimize problems of misclassification, such extension would also lead to diluting the standards themselves.  If we offer the same level of protection to everyone, we can offer much less than what those workers really in need of protection might need.  Imagine, for example, that the plumber or the tax consultant that you hire to help you occasionally will be treated as your employees, with the same rights that full-time employees of a large firm have via-a-vis that firm.  It is easy to see how this will lead to lower standards. In short, re-drawing the lines is necessary; avoiding them altogether is problematic.

Second, should we leave the determination of employee status to judges?

It is tempting to suggest that the question of “who is an employee” should be resolved with clear-cut answers in legislation or administrative regulations.  It is indeed useful to create more determinacy by allowing workers in specific sectors to know their status.  There have been some interventions of this kind in recent months with regard to Uber drivers, in both directions.  However a proclamation that specific workers will be deemed employees (or independent contractors) notwithstanding any other result reached by the regular tests, cannot replace the tests themselves.  This is an area in which some degree of indeterminacy is necessary.  New forms of work appear all the time.  If we set in legislation a specific list of criteria for clear-cut determination, it will be easy for employers to work around them and evade the law.  To prevent evasion as much as possible and provide solutions for new work arrangements, it is necessary to leave a wide margin of discretion for courts and labor boards.

Third, should we maintain an “all or nothing” dichotomy?

The distinction between employees and independent contractors has been described as a “binary divide”.  In recent years it is becoming increasingly clear that such a crude dichotomy is insufficient to deal with the multitude of work arrangements and their nuances.  A useful way to explain this is by reference to the spectrum of possibilities between universalism and selectivity.  A universal arrangement is one that applies to everyone; a selective arrangement is targeted only at specific groups.  There are obvious advantages to a universal approach: most notably, it is much easier for people to know their rights (and for society to enforce them) when all employees have the same rights.  In contrast, by using selective laws we can target a solution to a specific group (for example, a specific sector) that is more suitable for that group, sometimes at a higher or lower level of protection.

The challenge is to find the right balance between universalism and selectivity.  One useful method towards an optimal balance is to add a third (intermediate) group between employees and independent contractors.  One additional group should not make it too difficult for people to know their rights.  At the same time, it will allow us to bring into the scope of protection workers who share only some of the characteristics of employees, and should be covered by some (but not all) labor and employment laws.  Indeed, an intermediate group of “dependent contractors” (or some other name) already exists in many countries and its acceptance around the world is proliferating.  Although I disagree with the view that Uber drivers should be classified as being in such an intermediate group – for reasons that will be explained in the following posts – I agree entirely that the addition of such a group in the U.S. is warranted.

Fourth, how should we interpret the term “employee”?

Notwithstanding the unfortunate exception of the U.S., there is growing consensus among courts in other countries (such as Canada and the United Kingdom) that the term should be interpreted purposively: a term in legislation should be given a meaning that will best advance the purpose of that legislation.  Among academics this seems to be widely accepted (in principle) in the U.S. as well. So to decide who should be considered an “employee” we have to ask what are the goals (justifications) behind a specific law, or the goals behind the “project” of labor and employment law as a whole (because at some level, all labor and employment laws share the same basic purpose).  As part of the need to balance between universalism and selectivity, I believe we should aim for a general understanding of “who is an employee”, based on the project as a whole.  Then we can also make adjustments, when necessary, in light of other/more specific goals of specific legislation.

Adopting a purposive approach will also allow us to expose problems that cannot be resolved by interpretation and require legislative amendments.  For example, there are some specific protections for which the exclusion of independent contractors could be impossible to explain (i.e. for which the distinction as a whole seems unsuitable).  It may also be the case that new forms of work require new legal protections.  For example, to the extent benefits are unnecessarily tied to a specific workplace and accumulate over time, this is becoming increasingly detrimental for people who work for short periods or for multiple employers.  In other words, I agree entirely that new forms of work such as “on demand” work should trigger rethinking of existing laws and their suitability.  Nonetheless, as far as the employee/independent contractor distinction in itself is concerned, a purposive approach ensures its suitability for “on-demand” workers: when we encounter new work arrangements such as these, we should not ask whether they are similar to “traditional” employment or not, but rather whether this is the kind of arrangement that requires the application of labor and employment laws, given their purposes.  In other words, if Uber drivers are in need of labor and employment protection, purposive interpretation would ensure that they get it (assuming this is not explicitly prevented by legislation).

In the next post I will discuss briefly the goals of labor and employment laws and which tests/indicia can be derived from them for determining employee status.  For a much elaborated discussion of the issues covered in the first two posts please see my recent book.  In the third and final instalment I will apply these conclusions to the case of Uber drivers (whom I do not discuss specifically in the book).