Amazon's Dual Role Isn't the Antitrust Smoking Gun Regulators Assumed

Based on the research ofSharmistha Sikdar, Vrinda Kadiyali, and Giles Hooker, "Does Amazon's Dual Role Weaken Marketplace Competition?," Journal of Marketing, 2026

Regulators have spent three years building a case on the opposite assumption. A company that owns the shelf and also sells from it should be able to tilt the shelf in its own favor: rank its products higher, win the prime placement by default, and squeeze out the independent merchants who depend on it for distribution. The Federal Trade Commission and the European Union have each built major enforcement actions on versions of this logic, and the mechanism is intuitive enough that it has become the default framing for how hybrid marketplaces are discussed. A new empirical study by Sharmistha Sikdar, Vrinda Kadiyali, and Giles Hooker, published in the Journal of Marketing, tests that logic directly against Amazon's actual pricing and Buy Box behavior, and the results cut against the self-preferencing story on every mechanism they examine.

Three Mechanisms, One Counterintuitive Pattern

The researchers isolate three channels through which Amazon's dual role as marketplace owner and first-party (1P) seller could weaken competition for third-party (3P) sellers: price dependencies between 1P and 3P listings, the Buy Box selection process that determines which offer a shopper actually sees first, and barriers that might discourage new 3P sellers from entering a category Amazon also sells in.

On price dependencies, the expected story is that Amazon uses its visibility into 3P pricing to extract more from shoppers once rivals raise prices, either by matching them upward or by quietly tolerating a higher floor. The paper finds the reverse. Amazon's 1P prices generally drop in response to a high Buy Box price or to a significant price increase from 3P sellers. Rather than letting a 3P price spike become a signal to raise its own price in step, Amazon's pricing engine treats that spike as an opening to undercut it. The dual role, in this mechanism, behaves like a built-in price-disciplining device rather than a coordination tool, and the paper's authors describe it directly as evidence of downward discipline, not upward propping.

The second mechanism, the Buy Box, is the part of Amazon's marketplace that decides which seller's offer shoppers actually see highlighted and can add to cart with one tap. If self-preferencing were operating here, Amazon's own listings would hold an advantage that outside sellers cannot match regardless of what they charge. The paper finds instead that the Buy Box algorithm penalizes higher prices broadly, and that it also disadvantages low-reputation sellers independent of price. The box does not go to Amazon because Amazon owns the platform; it goes to whichever eligible offer, Amazon's own or a third party's, scores best on the blend of price and seller reliability the algorithm weighs. That matches what sellers who manage Buy Box eligibility for a living describe: price is one factor among several, with fulfillment method and seller performance metrics, among them order defect rate and shipment timeliness, carrying real weight in the ranking before price settles the remainder. Independent estimates put the Buy Box's importance at roughly 82 percent of Amazon sales flowing through whichever offer wins it at a given moment, which is precisely why its scoring logic is the mechanism antitrust scrutiny has focused on.

A Case and a Law Built on the Opposite Premise

The paper's findings land directly on top of live enforcement. On September 26, 2023, the FTC and seventeen state attorneys general sued Amazon in federal court, alleging the company illegally maintains monopoly power through a set of interlocking practices that include pressuring sellers toward Amazon's own fulfillment and advertising services and degrading the shopping experience in ways that protect its position. Separately, the European Commission designated Amazon, alongside five other firms, as a gatekeeper under the Digital Markets Act on September 6, 2023, with Amazon's Marketplace service specifically named. That designation subjects Amazon's marketplace operations to the DMA's rules against favoring its own products and services over those of competitors that depend on its platform to reach customers. Both actions assume a mechanism this paper tests and does not find: that Amazon's simultaneous roles as seller and shelf-owner systematically tilt outcomes toward its own listings and away from the merchants who sell alongside it.

A company that owns the shelf and sells from it is assumed to tilt the shelf. The pricing and entry data say this shelf tilts toward the customer instead.

None of this means the underlying antitrust concern is manufactured from nothing. Self-preferencing is a real, documented pattern elsewhere in platform markets, and the authors note that other researchers have found empirical support for Amazon favoring its own products in search rankings specifically. What this paper adds is a test of whether that tilt, where it exists, actually degrades competitive outcomes for 3P sellers once you look at prices and entry rather than ranking position alone. The answer, on their data, is no on both counts examined here.

The Entry Test Regulators Didn't Run

The third mechanism is the one with the least intuitive result and the most direct relevance to the barriers-to-entry question that sits at the heart of the FTC's theory of harm. If Amazon's dual role erects a disguised barrier, raising its own prices should be a signal to potential 3P entrants that the category is no longer worth entering, since Amazon can presumably still win the Buy Box or outcompete them on terms they cannot match. The paper finds higher 1P prices are correlated with increased 3P seller entry, not decreased entry. A rising Amazon price in a category reads, in this data, as an invitation rather than a warning: it signals room for a competitively priced outsider to undercut the incumbent and win share, exactly the dynamic a healthy hybrid marketplace is supposed to produce.

Put the three findings together and the paper arrives at a conclusion that runs against the prevailing antitrust narrative on Amazon specifically: despite real self-preferencing concerns raised by the FTC and the EU, the empirical evidence from pricing, Buy Box selection, and entry patterns suggests Amazon's dual role does not weaken marketplace competition, and may even support a more competitive hybrid marketplace through lower prices and open entry. That is a narrower claim than "Amazon is innocent of everything alleged." It is a claim about three specific, measurable mechanisms, tested against actual pricing and entry data rather than inferred from Amazon's incentives or its ranking algorithm's existence. The gap between what a dual-role platform could in theory do to competitors and what Amazon's own prices and seller entry actually show it doing is the gap this paper is built to close, and on these three tests, it closes in Amazon's favor.

For anyone running or regulating a hybrid marketplace, the lesson is not that self-preferencing concerns are overblown everywhere. It is that the specific mechanisms matter more than the structural fact of dual ownership, and that a platform's price response to competitors and its effect on entry are testable questions with answers that can run opposite to what the ownership structure alone would predict.