The Department of Labor Is Right to Dump Algorithmic Management

The phrase “algorithmic management” is applied to everything from scheduling software to chatbots to artificial intelligence.

Critics have marketed the term as something new and bad for American workers. According to academic journals, government reports, and the media, companies are controlling workers through an array of “digital tricks” where they record our keystrokes, track our locations, and even watch us through our webcams.

The research and stories on algorithmic management have made their way into federal regulations—specifically, in the Department of Labor’s current rule about independent contractors. This rule assumes that algorithmic management is pervasive, and it treats the practice a form of “control.”

But there’s a big problem: Algorithmic management isn’t a real thing. It is a term used to describe business practices so old and so normal that they would be boring if they weren’t branded with a scary label.

Fortunately, the narrative may finally run its course. DOL recently proposed eliminating the idea from its regulators and returning to a more traditional definition of workplace “control.” That would be a welcome development.

A Null Set

The term algorithmic management was coined in a 2015 academic paper focused on then-novel rideshare platforms. These platforms used digital tools to coordinate a vast fleet of independent drivers. The paper focused on three of these tools: matching riders to drivers, customer-ratings systems, and surge pricing. These tools not only helped platforms balance driver supply and rider demand, but also kept the drivers mostly happy. Algorithmic management, it seemed, wasn’t such a bad thing.

But since then, the term has taken on a life of its own. People such as Noam Scheiber of The New York Times, Samuel Levine, formerly of the Federal Trade Commission, and Veena Dubal of the University of California, Irvine School of Law have described algorithmic management in pernicious terms, arguing that companies are using it to fool people into working longer hours for less money.

These arguments have shaped public policy. In 2024, DOL adopted a regulation about the difference between employees and independent contractors. That difference is important because under the Fair Labor Standards Act, only employees are entitled to minimum wages and overtime. In part, the difference has always depended on who controls the work—the worker, or the person who hires the worker. The 2024 regulation defined this type of “control” to include some kinds of algorithmic management.

In particular, it pointed to “technological” monitoring, such as GPS tracking. That kind of monitoring, the regulation said, counted as control even if it wasn’t paired with any affirmative directions, instructions, or limits. Technological monitoring itself was a way to control another person’s work.

Recoding Control

That conclusion was a left turn from the DOL’s traditional rules. In effect, it treated some businesses’ practices as control even if they wouldn’t have been seen that way in a pre-digital world.

Consider monitoring itself. Before 2024, monitoring was a common and unobjectionable feature of business contracting. When a business contracts out work, it obviously wants to know that the work is getting done. But under the 2024 regulation, mere monitoring was considered control—at least when done through “technology.”

Or consider performance incentives. Before 2024, many business contracts included incentives for good service. A general contractor might pay extra to a subcontractor who finished a job early. But by the 2024 regulation’s logic, the extra payment might count as “control.” To everyone’s surprise, the subcontractor might become the general contractor’s employee—at least if the incentive was offered through a digital tool.

A Digital World

This kind of thinking gets the issue backward. Incentives, monitoring, and similar techniques aren’t signs of control: they’re signs of its absence. In economic terms, they’re just ways that businesses solve the “principle-agent problem.”

The basic issue is this: When a business hires a contractor, the two have different incentives. While the business wants the best service for the lowest price, the contractor wants the highest price for the least effort.

To protect its interests, the business must adopt safeguards against slack work. If it were using its own employees, it would manage that risk by telling the employees what to do. But it can’t do that with a contractor, often because it doesn’t know enough details about the work itself. Instead, it might require the contractor to report in when it hits certain milestones (monitoring). Or it might give the contractor an extra payment for high-quality work (incentives).

None of that changes with modern technology. Even when businesses and contractors use technology, they still face principle-agent problems. And the best tools for closing that gap are still monitoring, incentives, and similar methods. While these methods may work faster in the digital world, the dynamics are the same—even they’re wrapped in a scary label such as algorithmic management.

Fortunately, DOL is coming around to that view. In February, it published a new rule to distinguish between contractors and employees. This new rule says nothing about algorithmic management or any other kind of “technological” control.

Instead, it returns to first principles: It says that if a person controls her own work, she’s probably a contractor. If she doesn’t, she’s probably an employee. The point is the same even when work moves into the digital world.

For businesses, that means more certainty. They will no longer have to worry about special rules for technological tools. They can rely on the same principles they’ve relied on for years. Control is what it sounds like: it is instructions, requirements, restrictions. It is not incentives, progress reports, or data—“algorithmic” or not.

This article does not necessarily reflect the opinion of Bloomberg Industry Group, Inc., the publisher of Bloomberg Law, Bloomberg Tax, and Bloomberg Government, or its owners.

Author Guidelines

Alex MacDonald is an attorney in private practice in Washington, DC.

Tammy McCutchen contributed to this article.

Interested in writing? Review our author guidelines, and submit pitches to Insights@bloombergindustry.com.