The Journal of Things We Like (Lots)
Select Page
Omer Y. Pelled, The Proportional Internalization Principle in Private Law, 11 J. Legal Analysis 161 (2019).

One of the most challenging questions for present day tort law is who should be liable when an Uber car crashes and a passenger or a pedestrian is hurt, the driver or the platform? Similar legal dilemmas arise all over the platform economy. When a defective product sold by a vendor through Amazon’s Marketplace malfunctions or causes personal injuries, can the platform be held liable as a “seller”?

An entirely separate question has for long haunted contract law: should a party be held liable for abandoning negotiations prior to the formation of a contract, if it can be shown that the other party sunk non-salvageable investment in the course of the negotiations? Most courts say no, but in some notable exceptions courts have awarded the disappointed party it full reliance costs. Is this the right result?

These two seemingly unrelated puzzles have recently received a unified and persuasive theoretical treatment in Omer Pelled’s excellent article, The Proportional Internalization Principle in Private Law. Pelled argues that one underlying principle ought to shape the answer to these problems. He regards these as two illustrations of a general problem arising in multi-party interactions: How to apportion liability when the actions of one party, which caused the loss, benefitted others. The principle Pelled uncovers—“proportional internalization”—works by ensuring that each party internalizes an identical proportion of the costs and benefits.

In the Uber example, the driver’s efforts create revenue divided according to a fixed proportion with Uber (say, 70:30). The proportional internalization principle dictates that the cost of liability should also be borne by the parties in the same ratio. When an accident occurs, Uber should be liable for 30% of the harm. Not for the entire loss, as it might be held under a vicarious liability regime, nor a free pass. Dividing the costs of liability in the same proportion the benefits are shared would lead the driver to exert optimal effort.

Or, take the Amazon example. The question it poses has enormous stakes—whether the platform should be held liable for injuries from products sold by one of the many Marketplace vendors. The answer has split American courts. Technically, Amazon is not the “seller.” It doesn’t have title to the products sold by other vendors, and it merely provides the service to connect consumers with vendors. Many courts have therefore held that Amazon is not liable under products liability law (see, for example, a recent decision by a Federal Court in Arizona). But some new decisions, most prominently by the Third Circuit in Oberdorf v. Amazon, reached the opposite result.

The proportional internalization principle breaks the all-or-nothing dichotomy that seems to characterize court decisions. Rather than choosing between zero or full liability, Amazon ought to be liable in part, and its share should equal the share of the benefit it receives from the sale of the product over its platform. “Comparative benefit”—rather than comparative fault—should shape the allocation of liability, and provide optimal incentives for precaution to the vendors and the platform.

In contract law, the question how to share the benefits from a contract, especially when such benefits arise from an investment by one side, is determined by the contract, not by the law. But the law does affect the sharing rule when the contract breaks down. Imagine that the possibility of reaching an agreement leads one party (at the encouragement of the other) to make costly precontractual investments. These investments would be beneficial to both if an agreement is reached. Under the proportional internalization principle, if an agreement fails, the same one-party investments should be costly to both, divided in a manner that approximates how the benefits would have been split.

To illustrate, consider the investments a franchisee makes while negotiating a possible franchise with the corporation. The hopeful franchisee incurs costly reliance expenditures in relocating, training, buying property, and forfeiting other income opportunities, all under some assurance that negotiations are moving forward. If the negotiations fail, courts typically refuse to award the disappointed franchisee any damages: no contract means no liability. But sometimes courts take the opposite view, as the Wisconsin Supreme Court did in Hoffman v. Red Owl Stores, awarding Mr. Hoffmann damages for the non-salvageable portions of his reliance investment.

Again, the proportional internalization principle comes to the rescue. If the goal is to induce negotiating parties to make optimal reliance investments, liability for the investment costs should not be all or nothing. Instead, it should be apportioned based on the parties’ prospective shares in the expected surplus.

Pelled’s proportional internalization principle offers a broad-reaching unified treatment of many problems in private law and indeed the article is rich in illustrations from every corner of the law. It makes a crisp and fundamental observation about the design of optimal liability in situations of joint care or joint benefit. It also lends itself to endless applications that are not analyzed in the article, like the Amazon liability question. Of course, the basic intuition it builds on was recognized and even formally applied in specific contexts in prior literature, but the cross-substantive template was never elicited in such useful way. The article will have much value by offering a simple and intuitive mechanism to address many questions, old and new, in private law.

Download PDF
Cite as: Omri Ben-Shahar, Who Should Be Liable When Uber Cars Crash?, JOTWELL (June 1, 2020) (reviewing Omer Y. Pelled, The Proportional Internalization Principle in Private Law, 11 J. Legal Analysis 161 (2019)),