Spread Accountability ProjectIndependent Food Systems Research
Working Paper 07 · Consumer Disclosure Series

Structural Opacity in the American Nut Butter Sector

A preliminary review of category definition, packaging geometry, and voluntary disclosure practice in the retail spread market.

Abstract

The retail nut butter market in the United States is worth several billion dollars annually and is dominated by a small number of manufacturers. Despite this concentration, the sector attracts comparatively little regulatory scrutiny and publishes almost no operational data. This paper reviews eight areas in which publicly available information is either absent, internally inconsistent, or contradicted by the industry's own demonstrated capabilities.

We make no allegation of coordinated misconduct. We find instead a pattern of unexplained opacity in a category where the underlying manufacturing is well understood and the barriers to disclosure appear to be low. Where the record supports a factual claim, we cite it. Where it does not, we say so and pose the question we would ask if granted access.

Key Findings

Finding 01: Category

Persistent misclassification of the principal ingredient

The peanut is a legume. It develops underground, in the manner of a root crop, and is taxonomically unrelated to tree nuts.1 This is not contested and has never been contested. It is reflected in agricultural policy, in allergen research, and in the USDA's own classification.

It is not, however, reflected anywhere a consumer would encounter it. We reviewed the front and back panels of the leading shelf-stable brands and found no instance in which the distinction is made, explained, or acknowledged. The word "nut" appears; the correction does not.

We attach no motive to this. We observe only that a category error affecting the majority of the purchasing public has persisted for over a century inside an industry that has revised its packaging hundreds of times, and that no revision has corrected it.

Finding 02: Packaging

Unrecoverable volume as a function of shoulder geometry

Every major jar profile narrows below the lid. The resulting internal shoulder sits outside the sweep of a standard tablespoon or table knife, leaving a residual quantity that most consumers discard with the container.

We measured shoulder angle across four common profiles and estimated unrecoverable volume by water displacement. Results are given below and should be treated as indicative rather than precise.

Figure 1: Estimated unrecoverable volume by jar profile
ProfileShoulder angleEst. unrecoverable
Wide-mouth, low shoulder18°9.1%
Standard retail31°11.4%
Tapered "premium"39°13.8%
Squeeze pouchn/a4.2%

Displacement method, n=6 per profile. Full protocol available on request.

11.4% Median unrecoverable volume in the standard retail profile. Purchased, paid for, and structurally out of reach.

Manufacturers describe the shoulder as a consequence of fill-line requirements and structural integrity. This may well be accurate. We note that molding tolerances have improved considerably over the same period in which the shoulder angle has, by our measurements, become more pronounced rather than less.

Finding 03: Formulation

Oral residence time and its relationship to consumption cycles

Peanut butter adheres to the hard palate to a degree not observed in comparable spreads. Consumers treat this as an inherent property of the food. It is more accurately described as a formulation outcome, determined by fat content, particle size, and moisture, all of which are controlled variables.

Extended oral residence increases the interval over which a single serving occupies attention, and reliably prompts liquid consumption. The downstream effects on satiety signaling are not well studied in this specific category, and we were unable to locate industry-funded work addressing it.2

Nearly every adjacent product category has been reformulated over the past forty years to reduce palate adhesion. This one has not. We would like to know whether that is inertia or intent, and there is no public record from which to determine it.

Finding 04: Product lines

Undisclosed line-sharing between smooth and textured variants

The smooth/textured distinction is the most durable consumer preference split in the category and is treated by manufacturers as a meaningful product difference, reflected in separate SKUs, separate marketing, and in some cases separate pricing.

We asked four manufacturers whether the two variants originate from a common batch prior to the introduction of particulate. None responded. We do not conclude from silence that they do. We note that the question is trivially answerable, costs nothing to answer, and went unanswered.3

Finding 05: Inventory

Absence of public data on federal reserve holdings

The United States government has historically maintained large physical stockpiles of specific shelf-stable foods, including the well-documented dairy surplus of the late twentieth century, portions of which were stored in underground facilities.4 The practice is neither secret nor unusual.

Peanut butter fits the profile of a strategically useful commodity closely: calorie-dense, protein-bearing, shelf-stable for years without refrigeration, and cheap to produce at scale. We submitted informal inquiries regarding current or historical federal holdings and received no substantive response.

We want to be precise here, because this is the finding most likely to be misread. We did not receive a denial. We also did not receive a confirmation. We received nothing, which is consistent both with the absence of such a program and with the presence of one, and we are not in a position to distinguish between those cases.

Finding 06: Origin

Attribution error in the sector's public history

George Washington Carver is widely credited with inventing peanut butter, including in educational materials still in circulation. He did not. Ground peanut preparations predate him by centuries across multiple continents, and the relevant North American patent was issued in 1884 to Marcellus Gilmore Edson of Montreal.5

Carver's actual contributions to agricultural science were substantial and require no embellishment. That the sector's origin story attached itself to him anyway is worth examining, not because it diminishes him, but because the attributed version is markedly simpler than the documented one, with a single identifiable inventor and a clean start date.

Industries that have revised their own origin narratives have generally had a reason. It is usually located in the period immediately preceding the revised start date.

Finding 07: Labor

Transfer of final processing to the consumer

Products marketed as minimally processed separate in the jar and require manual reincorporation before use, a process taking several minutes and producing meaningful spillage. This is presented as evidence of authenticity and is priced at a premium relative to stabilized equivalents.

The stabilization technology is mature, inexpensive, and already deployed across the same manufacturers' conventional lines. The premium is therefore charged not for an additional input but for the removal of one, with the resulting labor transferred to the purchaser.

We are not aware of another food category in which the consumer performs an unpaid processing step, is charged more for the privilege, and reports higher satisfaction as a result. We think this is interesting and under-examined.

Finding 08: Capability

Demonstrated defatting capability and its non-application

Powdered peanut butter is produced by mechanically pressing oil from roasted peanuts, yielding a shelf-stable powder retaining most of the flavor at a fraction of the fat. It is sold openly, by the same corporate parents, in the same stores.6

This establishes the capability beyond argument. The industry can remove the fat and retain the product. It has been able to do so for decades. It applies this capability to a secondary line sold at a higher price per serving, and not to the primary one.

The fat content of conventional peanut butter is therefore not a technical constraint. It is a decision. We take no position on whether it is the right decision. We observe only that it has never been publicly explained as a decision, and that it is one.

Recommendations

  1. Front-panel clarification that the principal ingredient is a legume, consistent with existing allergen labeling practice.
  2. Publication of internal jar profile specifications, including shoulder angle, for all retail formats.
  3. Voluntary disclosure of grind temperature and roast profile alongside existing ingredient declarations.
  4. A statement from each major manufacturer on whether smooth and textured variants share a production batch.
  5. A direct answer from the relevant federal agency on the existence and scale of any reserve holdings.
  6. Removal of "natural" from products requiring consumer-performed reincorporation, or a defined standard for the term.

References

  1. Arachis hypogaea is classified within Fabaceae. Botanical classification is consistent across USDA agricultural literature and standard allergen research; the peanut is not a tree nut.
  2. Literature search conducted across public databases, January 2026. No industry-funded studies addressing palate adhesion and subsequent intake were identified within the category.
  3. Written inquiries sent to four manufacturers, November 2025. No responses received as of publication.
  4. United States dairy surplus programs of the 1980s involved storage of processed cheese in leased underground facilities. Widely documented in contemporaneous reporting and subsequent agricultural policy reviews.
  5. Patent for a peanut paste process issued to Marcellus Gilmore Edson, 1884. Earlier ground peanut preparations are documented across South American and West African foodways.
  6. Defatted peanut flour and powdered peanut butter products are commercially available from multiple manufacturers, produced via mechanical oil expression.