Essay · November 15, 2026 · 4,500 words · about 20 minutes
Two Months' Salary
How the price of a diamond gets made, and what to ask before you pay
This is how the price of a diamond gets made, written for anyone at a jewelry counter about to spend more than they planned, and I say how sure I can be of each piece. I am not a gemologist or a financial adviser, and nothing here is advice for your situation. Every price carries its date and will go stale.
Words worth knowing: 15 plain-English explanations
The words this essay leans on, each explained twice: first in plain English, then in one line for the expert. In the essay, a word with a dotted underline opens the same explanation.
- Carat
- How much a diamond weighs.
- For the expert a unit of mass for gems; dealer prices are quoted per carat.
- Comes up in: The legal record, and where things stand now
- 4Cs
- The four things a diamond gets graded on.
- For the expert color, clarity, cut and carat weight, the system GIA says it created; GIA's D-to-Z color scale dates from 1953.
- Comes up in: How a grade gets made
- Grading report
- A lab's written description of one stone.
- For the expert a record of graded characteristics; GIA says it "does not certify or appraise" stones and states no appraisal value. Check the number on the issuing lab's own site.
- Comes up in: How a grade gets made
- Appraisal
- Someone's written estimate of what a piece is worth.
- For the expert most are retail replacement valuations for insurance, per Jewelers of America; a fair market value appraisal usually comes in lower. Not a resale price.
- Comes up in: What a diamond resells for
- Laboratory-grown diamond
- A real diamond made in a machine instead of dug out of the ground.
- For the expert "essentially the same optical, physical, and chemical properties" as mined diamonds (FTC, 2018), grown by HPHT or CVD; the Guides call for a qualifier (16 CFR 23.12 and 23.25).
- Comes up in: How a grade gets made
- Simulant
- Something that looks like a diamond and is not one.
- For the expert a different material, such as moissanite (silicon carbide); the Guides call for the word "imitation" or "simulated" (16 CFR 23.25).
- Comes up in: What the seller is supposed to tell you
- Treatment
- Something done to a stone after it forms to make it look better.
- For the expert a process altering color or clarity; under 16 CFR 23.14 and 23.24 the Guides treat non-disclosure as deceptive if it is not permanent, needs special care or significantly affects value.
- Comes up in: What the seller is supposed to tell you
- Rough
- Diamonds as they come out of the ground, before cutting.
- For the expert mines and their sales arms sell rough; prices per carat mix every quality and are not comparable with polished retail prices.
- Comes up in: How a custom gets built
- Wholesale
- The price stores and dealers pay each other.
- For the expert the level a private seller sells into, which is why resale sits below retail.
- Comes up in: What a diamond resells for
- Keystone
- The shop's markup over what it paid.
- For the expert the trade's term for the markup over cost; Epstein reported 100 to 200 percent on a diamond and setting in 1982.
- Comes up in: What a diamond resells for
- Magic size
- A round weight, like half a carat or one carat, that people pay extra to reach.
- For the expert a weight threshold where price per carat jumps; Scott and Yelowitz (2010) measured it at 0.50 and 1.00 carat in pre-2010 online listings.
- Comes up in: Magic sizes
- Conflict diamonds
- Diamonds that paid for a rebel war.
- For the expert the Kimberley Process definition, limited to rebel movements; it does not cover state or corporate abuses, labor or environment.
- Comes up in: The industry's best case
- Reference price
- The crossed-out "was" price next to a sale price.
- For the expert a comparison price; misrepresenting price falls under the Guides' general rule (16 CFR 23.1).
- Comes up in: What the seller is supposed to tell you
- Signal
- Proof you mean something, because it cost you.
- For the expert a costly action that reveals intent (Camerer, 1988, on gifts; after Spence).
- Comes up in: The industry's best case
- Commitments decision
- A regulator closing a case in exchange for promises.
- For the expert an EU competition decision making offered commitments binding, with no requirement to find an infringement, as with De Beers and Alrosa (2006, upheld 2010).
- Comes up in: The legal record, and where things stand now
I bought a ring for my wife, and it sent me down a rabbit hole. This essay is what I found there.
The line with no first ad
Many people have heard that an engagement ring should cost two months' salary. I tried to find out where that came from. The best-known account of how diamonds were sold to Americans is Edward Jay Epstein's 1982 article in The Atlantic, "Have You Ever Tried to Sell a Diamond?" I searched two unofficial copies of it, one a translation, and found no salary rule in either; I could not open The Atlantic's own page.
The BBC reported in 2014 that De Beers advertising first suggested one month's salary, that in the United States it later became two, and that Japanese men were urged to spend three. The press does not agree on when: the BBC and Al Jazeera say the 1980s, and JCK's news director wrote in 2009 that it was the 1990s. The tagline is quoted two ways, as "How can you make two months' salary last forever?" and "How else can two months salary last forever?" I could not find the first ad. So the number many buyers carry into the store came, by the best accounts I have, from a seller's advertising.
How a custom gets built
For most of the twentieth century one company sold most of the world's diamonds. In a 2011 opinion, a US appeals court described the plaintiffs' allegation that De Beers held about 65 percent of the diamond market, and noted an earlier ruling's conclusion that its share had fallen to about 45 percent by 2006. Epstein, a journalist, wrote in 1982 that De Beers had proved to be "the most successful cartel arrangement in the annals of modern commerce." That is his word, not a court's.
Epstein reports that a copywriter at the agency N. W. Ayer wrote "A Diamond Is Forever" and that within a year it was De Beers' official motto. Al Jazeera dates the line to 1947, and reports that the share of American brides with a diamond ring rose from 10 percent in 1940 to about 80 percent in 1980.
There is a serious competing account. The legal scholar Margaret Brinig studied US demand for engagement rings from 1930 to 1985 "not as a Madison Avenue success story, but rather as a natural outgrowth of economic processes." She argued that demand rose after states abolished the right to sue over a broken engagement, and that the ring came to work as a kind of bond. I think both can be true: the need for a pledge was real, and advertising helped decide that the pledge would be a diamond and roughly what it should cost.
The legal record, and where things stand now
In 2004 De Beers Centenary AG pleaded guilty in federal court in Columbus, Ohio, to conspiring to fix the prices of industrial diamonds in 1991 and 1992, and was fined $10 million, according to the US Department of Justice. The plea was about industrial diamonds, not gems or rings.
Separately, De Beers agreed in 2005 and 2006 to pay $295 million to settle US class actions over gem diamonds; a federal court approved the deal in 2008 and an appeals court upheld it in 2011. About $135 million of it was set aside for consumers who bought diamonds or diamond jewelry for their own use. The claims were never tried. In Europe, the Commission in 2006 accepted binding commitments from De Beers to stop buying rough diamonds from the Russian miner Alrosa from 2009, and the EU's Court of Justice upheld that decision in 2010. That was a , which, the Court noted, requires no finding of an infringement.
That world is mostly gone. De Beers produced 24.7 million in 2024 and 21.7 million in 2025, according to its parent, Anglo American. World production in 2025 was 98.8 million carats, according to the Kimberley Process's own statistics. By my rough sum that puts De Beers at about a fifth of the world's carats by volume. Mining Weekly reported that De Beers lost $511 million in 2025 before interest, tax, depreciation and amortization (EBITDA), and that Anglo American wrote its holding down by $2.3 billion, a fall the trade paper linked in part to buyers shifting between natural and lab-grown stones. Anglo American is selling its stake. Botswana's president named a preferred buyer in November 2026, but as of November 13 no binding sale had been signed, Reuters and JCK reported.
So no single seller is in a position to hold the price up anymore. What outlived the old structure is the custom it helped build: the ring, the salary line, the round sizes, which I think now keep themselves going through the people who buy.
What a diamond resells for
I most wanted a number here and found none: no public, current dataset of what used engagement diamonds sell for compared with what was paid.
What exists is old. In a 1970 British test reported by Epstein, a consumer magazine bought two half-carat stones for £400, and more than eight years later the best offer was £500. In a 1976 Dutch test, 19 of 20 dealers would not buy the stone at all, and the twentieth "offered only a fraction of the purchase price." Those are two small tests, about fifty years old. The resale site Worthy wrote in 2023 that resale is "sometimes as low as 20%-30% of the retail price," and the post gives no sample or method.
Against that, Renneboog and Spaenjers found that rare white diamonds sold at auction returned 6.4 percent a year above inflation, in US dollars, from 1999 to 2010. It is a different market from a mall ring, and the data end in 2010.
Why a typical stone resells low is ordinary, and this part is my inference. When you sell, you sell to the trade, which can already buy from dealers at . What you paid included the store's costs. In 1982 Epstein reported that the , the trade's word for the markup, on a diamond and its setting could run from 100 to 200 percent. Today's best number comes from the largest US jewelry chain, Signet, which owns Kay, Zales and Jared: for fiscal 2026 it reported $6,813.6 million in sales and $2,694.6 million of gross margin, about 40 percent of sales by my arithmetic, across everything it sells. Its operating income was 5.8 percent of sales, so most of that margin paid for stores and staff, not profit. Signet comes up often here because it is the largest chain and its record is public, not because the record shows it is worse than its rivals.
One confusion sits on top of this. Most today are written for insurance, as an estimate of what it would cost to replace the piece at a jewelry store, according to Jewelers of America, the retailers' trade group; an appraisal of what a piece would actually sell for usually comes in lower. The appraiser may be the seller.
How a grade gets made
GIA describes itself as the creator of the : color, clarity, cut and carat weight. GIA says it "does not certify or appraise" stones, and its state no appraisal value. Color and clarity are graded by trained people under controlled conditions, so a grade is a professional judgment, not an instrument reading like weight.
In 2014 the trading network RapNet said it would stop listing diamonds graded by EGL USA and EGL International, two separate companies, citing "misrepresentation of diamond quality" and inconsistency among EGL reports, National Jeweler reported. Three buyers who sued a Nashville retailer over stones sold with EGL International reports settled on confidential terms, so no court ruled on the grading, National Jeweler reported. The lab's owners then said it would stop using the EGL International name. I found no published blind study sending the same stones to several labs, so I am not repeating the claim that one lab grades more softly than another.
In June 2025 GIA said it would stop grading new lab-grown stones on its natural-diamond color and clarity scale and would call them "premium" or "standard," because "More than 95% of entering the market fall into a very narrow range of color and clarity." Its older lab-grown reports remain valid. IGI, another large lab, said it would keep grading lab-grown stones on the same scale it uses for natural ones, Jewellery Monthly reported, so two reports on two stones may not compare.
Magic sizes
Scott and Yelowitz, using the listed prices of three large online retailers, found premiums "upward of 18%" for a stone of one-half carat rather than slightly less, and "between 5% and 10%" for one carat rather than slightly less. Their data predate 2010, so the sizes of the premiums are history. The pattern is the point: at half a carat and at one carat, reaching the round number costs more than the extra weight explains.
By my own arithmetic, a 0.90-carat round stone is 10 percent lighter than a 1.00. If the two are cut the same way, it is only about 3.5 percent narrower across the top, because width grows with the cube root of weight.
Lab-grown, and what happened to its price
Under the FTC's Jewelry Guides, a laboratory-grown diamond is a diamond. In 2018 the agency removed the word "natural" from its definition of diamond, because lab-created products with "essentially the same optical, physical, and chemical properties" as mined diamonds "are also diamonds." The Guides treat it as deceptive to sell a lab-grown stone under a gem's name unless a word such as "laboratory-grown" or "laboratory-created" comes immediately before it, just as prominently, and "faux" is not enough.
When De Beers launched its Lightbox brand in 2018, it priced lab-grown stones at $800 for one carat. Announcing Lightbox's closure in May 2025, it said lab-grown prices "have fallen 90% at wholesale" and that it expected them "to fall further." JCK reported in August 2026 that the analyst Edahn Golan's lab-grown wholesale index was "now down 96% since he began tracking prices in 2018." On November 14, 2026, StoneAlgo's listing averages were $3,964 for a one-carat natural stone and $527 for a one-carat lab-grown, averaged across mixed grades, so not a like-for-like pair.
Retail margins on lab-grown stay wide: JCK reported in 2026 that US retailers still mark lab-grown up "more than 80% above wholesale," citing Paul Zimnisky's data. De Beers mostly sells mined stones, though it ran Lightbox until 2025, so I read its statements on lab-grown value as informative but not neutral.
Mined stones have fallen too. Rapaport's index of trade asking prices for better-quality round stones fell 9.9 percent for one-carat stones and 26.0 percent for half-carat stones over 2025, while three-carat stones barely moved. "Natural holds its value" is a weaker line than it was. Meanwhile lab-grown has taken over much of the counter: The Knot's 2026 survey of its own members found 61 percent of engagement rings had a lab-grown center stone.
None of this is an ad for lab-grown. It has lost far more of its price than mined stones, retailers keep wide margins on it, and the same questions apply to both.
Paying over time
In 2019 Sterling Jewelers, Signet's unit that runs Kay and Jared, agreed to consent orders with the Consumer Financial Protection Bureau and New York's attorney general and paid $11 million in penalties. The CFPB said Sterling had opened store credit-card accounts without customers' consent, signed customers up for payment-protection insurance without their consent and misrepresented financing terms. Signet said it disagreed with the allegations and settled to avoid the cost of litigation.
On store cards generally, the CFPB explained in 2017 that under a deferred-interest promotion, if any balance remains when the promotion ends, "consumers are charged accrued interest on the promotional balance from the time of purchase." I do not have reliable current figures on what store financing costs.
What the seller is supposed to tell you
The FTC's Jewelry Guides are guidance on what the agency considers deceptive. They are not a statute, and falling short of one is not automatically illegal. Within that limit, they say:
- "Cultured" can be used for lab-grown diamonds only with a clear disclosure that the stone is not mined.
- A should be disclosed if it is not permanent, needs special care, or "has a significant effect on the stone's value."
- Misrepresenting a piece's price or value is, in the Guides' general rule, deceptive (16 CFR 23.1).
In 2019 FTC staff sent eight warning letters to marketers of lab-grown and stones, about ads that might imply a simulant is a diamond and about words like "eco-friendly" and "sustainable," for which it was "highly unlikely that they can substantiate all reasonable interpretations of these claims." Warning letters are not findings of a violation.
On sale prices, the trade term for the crossed-out "was" number is a . The Guides' general rule on misrepresenting price covers it, and question 6 below is the one to ask.
The industry's best case
Diamonds are around 80 percent of Botswana's exports and a quarter of its economy, according to the IMF as reported by AP. In 2024, the IMF reported, mining output fell 24 percent and the economy shrank 3 percent. My reading: every buyer who chooses a lab-grown stone over a mined one is, at the margin, part of the fall in demand that shrank that economy. The IMF itself calls for "a more diversified export base."
The Kimberley Process, in force since 2003, certifies rough diamonds against , which it defines as "rough diamonds used by rebel movements or their allies to finance conflict aimed at undermining legitimate governments." Global Witness left in 2011, saying the scheme "has done much that is useful but ultimately has failed to deliver." As I read it, that definition covers one kind of harm, not abuses by governments or companies, working conditions or the environment.
The US Department of Labor lists diamonds from six countries for child or forced labor, mostly places with small-scale digging, and does not list Botswana, Canada, Russia, South Africa or Namibia for diamonds. In Surat, India, where both mined and lab-grown stones are cut, the Diamond Workers Union Gujarat said in 2024 that wages had halved, as ThePrint reported. Falling demand for mined stones is part of that story, and it is no argument for either kind.
On the environment, the miners' trade body released a 2019 Trucost study finding lower emissions per carat for mined stones; JCK called the comparison tentative, and a lab-grown executive disputed it.
And people value a mined stone for reasons that are not foolish. The economist Colin Camerer argued that gifts serve as "" of a person's intentions about a relationship, and that "inefficient gifts can be better signals." A ring that cannot be resold for what it cost may be doing exactly its job. That is the strongest case for the custom, and I accept it, with one limit: a signal has to cost the giver something, not match a salary multiple from an ad. One survey by Andrew Francis-Tan and Hugo Mialon found men who spent $2,000 to $4,000 on a ring were "1.3 times more likely to get divorced" than men who spent $500 to $2,000, as KSL reported in 2014. That is a correlation in an online sample and shows no cause.
Nothing here says a diamond someone already owns was a mistake. A symbol is worth what it means to the two people who share it. I think people should buy a diamond because it is beautiful, and not because they were told it is valuable. My argument is only with the gap between what a buyer is told and what the record shows: "forever," "investment," an appraisal above the price paid.
Before you pay
None of this tells you what to buy. These are questions to ask, with the rule beside those that have one.
- Is this stone mined or laboratory-grown, and is that written on the receipt? (FTC Jewelry Guides, 16 CFR 23.12 and 23.25)
- Has it been treated? Is the treatment permanent, and does it need special care? (16 CFR 23.14 and 23.24)
- Which lab graded it, and can I check the report number on that lab's own site before I pay?
- What is the return policy in writing: how many days, cash or credit, any restocking fee?
- Will you buy it back, at what percentage of what I pay, in cash, in writing? What are the trade-up terms?
- Did this item actually sell at the "was" price, and for how long? (16 CFR 23.1)
- Is this appraisal a replacement estimate for insurance or a resale price, and who wrote it? (16 CFR 23.1)
- What backs any "eco-friendly" or "sustainable" claim?
- What does the financing cost in total, and am I being enrolled in a store card or insurance?
- Is there a stone just under the round weight? Ask for the price per carat of each.
I built a free checklist and calculator for these questions: https://knownquantityllc.com/for-good/diamonds/
The stone is very old. Most of what sets its price is not.
Sources and a note on method
History and custom: Edward Jay Epstein, "Have You Ever Tried to Sell a Diamond?", The Atlantic, February 1982, https://www.theatlantic.com/magazine/archive/1982/02/have-you-ever-tried-to-sell-a-diamond/304575/ (not opened; searched in two unofficial copies, one a Chinese translation). BBC News Magazine, Laurence Cawley, May 2014 (read through a repost; the original page was not found). Al Jazeera, "How the diamond engagement ring was invented and sold around the world," September 14, 2025, https://www.aljazeera.com/features/2025/9/14/how-the-diamond-engagement-ring-was-invented-and-sold-around-the-world . JCK, Rob Bates, "The Two-Month Salary Guideline," April 10, 2009, https://www.jckonline.com/editorial-article/the-two-month-salary-guideline/ . Margaret F. Brinig, "Rings and Promises," Journal of Law, Economics and Organization 6 (1990), https://scholarship.law.nd.edu/law_faculty_scholarship/576 (abstract only).
Legal record: US Department of Justice press release, July 13, 2004, https://www.justice.gov/archive/opa/pr/2004/July/04_at_476.htm (opened on the official site). Sullivan v. DB Investments, Third Circuit en banc opinion, December 20, 2011, read on a law firm's copy, https://www.mmwr.com/wp-content/uploads/2011/12/082784pen.pdf (not a court site). Court of Justice of the European Union, Commission v Alrosa, Case C-441/07 P, June 29, 2010, press material, https://curia.europa.eu/jcms/jcms/P_65646/en (opened on the official site).
The company today: Anglo American Q4 2025 Production Report, February 5, 2026, https://www.angloamerican.com/media/press-releases/2026/05-02-2026 . Kimberley Process, Annual Global Summary: 2025, https://kimberleyprocessstatistics.org/static/pdfs/public_statistics/2025/2025Charts.pdf (the scheme's own statistics site), and Rapaport News, July 1, 2026, https://rapaport.com/news/russia-largest-rough-producer-by-value-in-2025-for-third-year/ . Mining Weekly, February 20, 2026, https://www.miningweekly.com/article/de-beers-loss-widens-to-511m-amid-ongoing-diamond-market-challenges-2026-02-20 . JCK, Rob Bates, November 13, 2026, https://www.jckonline.com/editorial-article/botswanas-comments-sale/ . Reuters via Kitco, November 13, 2026, https://www.kitco.com/news/off-the-wire/2026-11-13/angola-seeks-10-stake-de-beers-under-anglos-proposed-exit . Rapaport News on Botswana's choice of preferred bidder, November 10, 2026, https://rapaport.com/news/botswana-confirms-gdc-as-de-beers-bidder-seeks-30-stake/ . AP, February 25, 2025, https://www.ksat.com/news/world/2025/02/25/botswana-the-country-with-the-biggest-diamonds-gets-an-improved-10-year-deal-with-miner-de-beers/ .
Prices, resale and appraisals: Signet Jewelers fiscal 2026 results release, March 19, 2026, https://s26.q4cdn.com/755441662/files/doc_news/Signet-Jewelers-Reports-Fourth-Quarter-and-Full-Year-Fiscal-2026-Results-2026.pdf . Renneboog and Spaenjers, "Hard assets: The returns on rare diamonds and gems," Finance Research Letters (2012), https://econpapers.repec.org/RePEc:eee:finlet:v:9:y:2012:i:4:p:220-230 (abstract only). Worthy blog, February 22, 2023, https://blog.worthy.com/knowledge-center/diamonds/engagement-ring-value-can-fluctuate-time/ . Jewelers of America, jewelry appraisers guide, https://jewelers.org/education/choose-a-jeweler/jewelry-appraisers . Rapaport press release, January 6, 2026, https://www.businesswire.com/news/home/20260106871951/en/Rapaport-Press-Release-Diamond-Market-Cautious-at-Start-of-Year . StoneAlgo Diamond Price Index, figures for November 14, 2026, read November 15, 2026, https://www.stonealgo.com/diamond-prices/ . JCK, August 10, 2026, on Edahn Golan's and Paul Zimnisky's data, https://www.jckonline.com/editorial-article/lab-grown-wholesale-freefall/ . De Beers Group Lightbox launch release, May 29, 2018, https://prnewswire.com/news-releases/de-beers-group-to-launch-new-fashion-jewelry-brand-with-laboratory-grown-diamonds-300655304.html , and closure announcement, May 2025, via a trade repost, https://benchpeg.com/news/press-release-de-beers-group-announces-intention-to-close-lightbox-business/ . The Knot Worldwide, 2026 Real Weddings Study, https://www.theknotww.com/press-releases/the-knot-worldwide-unveils-2026-real-weddings-study .
Grading: GIA, 4Cs pages, https://4cs.gia.edu/en-us/4cs-diamond-quality/ ; GIA FAQ on grading and appraisal, https://www.gia.edu/gia-faq-analysis-grading-certify-appraise-diamond-gemstone ; GIA, "New descriptive terminology for laboratory-grown diamonds," June 2, 2025, https://www.gia.edu/gia-news-press/new-descriptive-terminology-for-laboratory-grown-diamonds Jewellery Monthly, July 19, 2025, on IGI, https://jewellerymonthly.co.uk/igi-maintains-4cs-grading-for-lab-grown-diamonds-amid-industry-shift/ . National Jeweler on RapNet and EGL, September 9, 2014, https://nationaljeweler.com/articles/5920-rapnet-to-de-list-egl-graded-diamonds , and December 2014, https://nationaljeweler.com/articles/5923-remaining-lawsuits-filed-over-egl-reports-are-settled . Scott and Yelowitz, "Pricing Anomalies in the Market for Diamonds: Evidence of Conformist Behavior," Economic Inquiry (2010), https://ideas.repec.org/a/bla/ecinqu/v48y2010i2p353-368.html (abstract only).
Rules and enforcement: FTC Jewelry Guides, 16 CFR Part 23, sections 23.1, 23.12, 23.14, 23.24 and 23.25, https://www.ecfr.gov/current/title-16/chapter-I/subchapter-B/part-23 (opened on the official site). Federal Register, 83 FR 40665, August 16, 2018, https://www.federalregister.gov/documents/2018/08/16/2018-17454/guides-for-the-jewelry-precious-metals-and-pewter-industries (opened on the official site). FTC Business Blog, July 24, 2018, https://www.ftc.gov/business-guidance/blog/2018/07/diamonds-and-pearls-and-revisions-ftcs-jewelry-guides (opened on the official site). FTC press release on warning letters, April 2, 2019, https://www.ftc.gov/news-events/news/press-releases/2019/04/ftc-sends-warning-letters-companies-regarding-diamond-ad-disclosures (opened on the official site). CFPB enforcement action, Sterling Jewelers Inc., January 16, 2019, https://www.consumerfinance.gov/enforcement/actions/sterling-jewelers-inc/ (opened on the official site). National Jeweler, January 17, 2019, https://nationaljeweler.com/articles/4072-signet-jewelers-fined-11m-over-in-store-credit-cards . CFPB release on deferred-interest promotions, June 8, 2017, https://www.consumerfinance.gov/about-us/newsroom/consumer-financial-protection-bureau-encourages-retail-credit-card-companies-consider-more-transparent-promotions (opened on the official site).
Producing countries and the record: IMF Press Release 25/317, September 26, 2025, https://www.imf.org/en/news/articles/2025/09/26/pr-25317-botswana-imf-staff-completes-2025-article-iv-mission-to-botswana (opened on the official site). Kimberley Process, "What is the KP," https://www.kimberleyprocess.com/en/what-kp (the scheme's own site). Global Witness, Charmian Gooch, December 3, 2011, https://www.globalwitness.org/archive/why-we-are-leaving-kimberley-process-message-global-witness-founding-director-charmian-gooch/ . US Department of Labor, 2024 List of Goods Produced by Child Labor or Forced Labor, https://www.dol.gov/agencies/ilab/reports/child-labor/list-of-goods (opened on the official site). ThePrint, November 4, 2024, https://theprint.in/india/job-losses-factory-closures-pushing-surats-diamond-workers-to-the-edge-71-suicides-in-18-months/2339805 . JCK, May 2, 2019, on the Trucost study, https://www.jckonline.com/editorial-article/diamond-miners-eco-friendly-claims/ .
Ideas: Colin Camerer, "Gifts as Economic Signals and Social Symbols," American Journal of Sociology (1988), https://authors.library.caltech.edu/records/nskm8-qzq73 (abstract). Francis-Tan and Mialon on wedding spending and marriage length, as reported by KSL, October 10, 2014, https://www.ksl.com/article/31903509 (the journal article was not opened).
Every price and percentage above was current on the date given beside it and will go stale.
I used AI to research and draft this, under my direction. The first draft then went to a second review, also done with AI, which had not written the draft and read it with a defamation lawyer, a gemologist, an FTC staff attorney and a recent ring buyer in mind. No lawyer or gemologist has read this essay. The review checked 33 claims against the sources listed here, and reopened these on the official sites: the Justice Department's 2004 release, the FTC's Jewelry Guides on ecfr.gov and its 2019 warning-letter release, the CFPB's Sterling page and its 2017 deferred-interest release, the EU Court of Justice's press release and the IMF's Botswana release, along with the Kimberley Process's own page. It found a market-share figure I had presented as a court's when it was the plaintiffs' allegation and an earlier ruling's conclusion, a write-down attributed to the wrong company, two lawsuits against one company presented as a pattern across chains, a company reply with no reliable source, one width figure I computed the wrong way round, and about a dozen claims stated more strongly than their sources. Those are corrected here. On November 15, 2026, every claim and source link was checked again, also with AI. That recheck found a newer world production figure and new StoneAlgo prices, and those are updated here. It also found that one of the two private lawsuits over sale prices had been withdrawn without a ruling and that it could not see where the other stands now, so I took both out. It could not open The Atlantic's own copy of Epstein or the US settlement agreement, so this essay does not say whether De Beers admitted anything.
Disclosure: I bought a ring for my wife, which is what started this. Nothing in this essay is a recommendation to buy or sell anything.
Carat
How much a diamond weighs.
For the expert a unit of mass for gems; dealer prices are quoted per carat.
4Cs
The four things a diamond gets graded on.
For the expert color, clarity, cut and carat weight, the system GIA says it created; GIA's D-to-Z color scale dates from 1953.
Grading report
A lab's written description of one stone.
For the expert a record of graded characteristics; GIA says it "does not certify or appraise" stones and states no appraisal value. Check the number on the issuing lab's own site.
Appraisal
Someone's written estimate of what a piece is worth.
For the expert most are retail replacement valuations for insurance, per Jewelers of America; a fair market value appraisal usually comes in lower. Not a resale price.
Laboratory-grown diamond
A real diamond made in a machine instead of dug out of the ground.
For the expert "essentially the same optical, physical, and chemical properties" as mined diamonds (FTC, 2018), grown by HPHT or CVD; the Guides call for a qualifier (16 CFR 23.12 and 23.25).
Simulant
Something that looks like a diamond and is not one.
For the expert a different material, such as moissanite (silicon carbide); the Guides call for the word "imitation" or "simulated" (16 CFR 23.25).
Treatment
Something done to a stone after it forms to make it look better.
For the expert a process altering color or clarity; under 16 CFR 23.14 and 23.24 the Guides treat non-disclosure as deceptive if it is not permanent, needs special care or significantly affects value.
Rough
Diamonds as they come out of the ground, before cutting.
For the expert mines and their sales arms sell rough; prices per carat mix every quality and are not comparable with polished retail prices.
Wholesale
The price stores and dealers pay each other.
For the expert the level a private seller sells into, which is why resale sits below retail.
Keystone
The shop's markup over what it paid.
For the expert the trade's term for the markup over cost; Epstein reported 100 to 200 percent on a diamond and setting in 1982.
Conflict diamonds
Diamonds that paid for a rebel war.
For the expert the Kimberley Process definition, limited to rebel movements; it does not cover state or corporate abuses, labor or environment.
Reference price
The crossed-out "was" price next to a sale price.
For the expert a comparison price; misrepresenting price falls under the Guides' general rule (16 CFR 23.1).
Signal
Proof you mean something, because it cost you.
For the expert a costly action that reveals intent (Camerer, 1988, on gifts; after Spence).
Commitments decision
A regulator closing a case in exchange for promises.
For the expert an EU competition decision making offered commitments binding, with no requirement to find an infringement, as with De Beers and Alrosa (2006, upheld 2010).
Questions this essay answers
Where does the two months' salary rule come from?
The BBC reported in 2014 that De Beers advertising first suggested one month's salary, that in the United States it later became two, and that Japanese men were urged to spend three. I could not find the first ad. So the number many buyers carry into the store came, by the best accounts I have, from a seller's advertising.
Does De Beers still control diamond prices?
De Beers produced 24.7 million carats in 2024 and 21.7 million in 2025, according to its parent, Anglo American. World production in 2025 was 98.8 million carats, according to the Kimberley Process's own statistics. By my rough sum that puts De Beers at about a fifth of the world's carats by volume. So no single seller is in a position to hold the price up anymore.
What does a diamond resell for?
I most wanted a number here and found none: no public, current dataset of what used engagement diamonds sell for compared with what was paid. The resale site Worthy wrote in 2023 that resale is "sometimes as low as 20%-30% of the retail price," and the post gives no sample or method.
Is an appraisal what a ring would sell for?
Most appraisals today are written for insurance, as an estimate of what it would cost to replace the piece at a jewelry store, according to Jewelers of America, the retailers' trade group; an appraisal of what a piece would actually sell for usually comes in lower. The appraiser may be the seller.
Is a lab-grown diamond a real diamond?
Under the FTC's Jewelry Guides, a laboratory-grown diamond is a diamond. In 2018 the agency removed the word "natural" from its definition of diamond, because lab-created products with "essentially the same optical, physical, and chemical properties" as mined diamonds "are also diamonds."
Why does a one-carat stone cost more than a slightly smaller one?
Scott and Yelowitz, using the listed prices of three large online retailers, found premiums "upward of 18%" for a stone of one-half carat rather than slightly less, and "between 5% and 10%" for one carat rather than slightly less. Their data predate 2010, so the sizes of the premiums are history. The pattern is the point: at half a carat and at one carat, reaching the round number costs more than the extra weight explains.
What should I ask before I pay?
None of this tells you what to buy. These are questions to ask, with the rule beside those that have one. The essay lists ten, from whether the stone is mined or laboratory-grown to the price per carat of a stone just under the round weight. I built a free checklist and calculator for these questions.
