New York University Public Law and Legal Theory Working Papers

Document Type



Will advances in robotics, artificial intelligence, and machine learning put vast swaths of the labor force out of work or into fierce competition for the jobs that remain? Or, as in the past, will new jobs arise to absorb workers displaced by automation? These hotly debated questions have profound implications for labor and employment law, and for the fortress of social entitlements that has been built on the foundation of the employment relationship. That is because, first, the law of work effectively “taxes” the employment of human labor; to the extent that it adds to labor costs, it is a factor both in firms’ flight from direct employment through various forms of “fissuring,” and in their substitution of machines for human workers. Moreover, the prevailing legal responses to fissuring – which aim to extend firms’ legal responsibility for the workers whose labor they rely on – cannot address the distinctive challenge of automation, and might modestly exacerbate it. Automation offers firms the ultimate exit from the costs, risks, and hassles associated with human labor. As technology becomes an ever more capable and cost-effective competitor to human workers, it may doom the prevailing strategy of shoring up the fortress of employment.

In this article I propose a general strategy for reducing the legal tax on employment – and for marginally reducing firms’ incentive to replace employees with contractors and human workers with machines – while protecting the essential rights and entitlements of all of those who work for a living. The basic move is to separate the question of what workers’ entitlements should be from the question of where their economic burdens should fall. Some worker entitlements are inseparable from employer duties. But for those that are not, we should consider ways to extend the entitlements, their funding, or both so that we protect all those who need protection while reducing the legal tax on the use of human labor.

Date of Authorship for this Version

Summer 7-2017