Why $199 Feels Cheaper Than $200: Anchoring Data

A penny shouldn’t move markets. Yet when a price crosses from $2.99 to $3.00, consumers behave as if it rose by 15 to 25 cents — not one. That estimate comes from Avner Strulov-Shlain’s structural analysis of retail scanner data covering 3,500 products across 25 US supermarket chains, published in the Review of Economic Studies in 2023. The single penny carries the weight of a quarter because the leftmost digit changed from a 2 to a 3, and the brain anchors on that digit.

This is the engine behind every $199 price tag in your inbox. The mechanism is well-documented, the magnitude is now quantified, and for households spending at higher absolute price points, the dollar cost of falling for it scales with the size of the purchase. The data below separates what the left-digit effect actually does from the marketing folklore around “charm pricing.”

The penny that costs a quarter

Start with the key figures, because the entire argument rests on a small number of well-sourced numbers.

Sources as listed; figures drawn from each study’s published abstract and results.

The condition matters more than the penny. Thomas and Morwitz established in their 2005 Journal of Consumer Research paper, “Penny Wise and Pound Foolish,” that nine-ending prices read as meaningfully lower than a price one cent higher only when the leftmost digit changes. A move from $3.00 to $3.01 produces almost nothing. A move from $2.99 to $3.00 produces the full effect, because $2.99 gets filed in the “two-dollar” bin and $3.00 in the “three-dollar” bin. The cent is incidental; the digit category is everything.

Strulov-Shlain’s contribution was to put a coefficient on it. Analyzing how demand actually moved across thousands of real price points, he found consumers respond to that one-cent crossing as though the price had climbed 15 to 25 cents. The behavioral premium is roughly 15 to 25 times the nominal change. That figure anchors everything that follows, and it is the number most popular coverage of charm pricing never cites — most articles assert that $9.99 “feels cheaper” without quantifying by how much.

Where the effect came from: anchoring

The left-digit effect is a special case of a broader mechanism. Tversky and Kahneman named it in 1974 in Science: anchoring, the tendency to start an estimate from an initial value and adjust insufficiently away from it. The adjustment falls short even when the starting number is plainly irrelevant. A 99-ending price supplies the anchor — the lower digit category — and the mind under-adjusts upward.

How far short can adjustment fall? Ariely, Loewenstein and Prelec tested exactly that in their 2003 Quarterly Journal of Economics study. MIT students wrote down the last two digits of their Social Security number, then bid on products in a real auction with real money. Students whose Social Security digits landed in the top fifth paid up to 346% more for the same items than students whose digits landed in the bottom fifth. For a wireless keyboard, the top-fifth students paid about $56 on average and the bottom-fifth about $16. The ordering of preferences stayed rational — everyone preferred the better wine to the lesser one — but the absolute dollar figures were anchored to a number every participant knew was meaningless. This is the same cognitive shortcut that makes anchoring a load-bearing concept across pricing psychology, retail and beyond.

Anchoring is not confined to small tickets. It scales, which is precisely why it matters at higher price points and is worth understanding alongside related levers like the decoy effect in product menus and the conditions of loss aversion in premium purchases.

The effect survives in markets with real money on the table

Laboratory anchoring is easy to dismiss as an artifact of student volunteers. The field evidence is harder to wave away. Anderson and Simester ran three randomized field experiments in women’s apparel catalogs, published in 2003 in Quantitative Marketing and Economics. Use of a $9 price ending increased demand in all three experiments, and the increase was stronger for new items than for items the retailer had sold in previous years. A garment priced at $39 outsold the same garment at $34 in their data — demand rose despite the higher price, because the left digit read as a discount signal.

The starkest evidence comes from a market where the stakes are large and the information is sitting in plain view. Lacetera, Pope and Sydnor analyzed over 22 million wholesale used-car transactions for their 2012 American Economic Review paper. Sale prices dropped discontinuously at 10,000-mile odometer thresholds, with smaller drops at 1,000-mile thresholds. A car reading 79,900 miles sold for roughly $150 to $210 more than an otherwise identical car reading just over 80,000 — for a difference of a few hundred miles. The buyers were focused on the leftmost odometer digit and partially ignoring the rest, in a transaction worth thousands of dollars. The pattern was driven by final customers rather than professional dealers.

Two markets, two product classes, two decades apart, and the same digit-anchoring shows up in actual purchase behavior with money committed. The effect is not a quirk of how people answer survey questions.

Finluxy Price-to-Quality Ratio applied to the 99-ending trap

The left-digit effect doesn’t change a product’s quality — it changes the perceived price. To make that distortion legible, the Finluxy Price-to-Quality Ratio compares an item’s quality score against the category median, divided by its price against the category median. A ratio above 1.0 means better value than the median; below 1.0 means you are paying more, relative to quality, than the median buyer.

The ratio’s value here is diagnostic: a 99-ending price and a round price for the same physical good deliver identical quality but are perceived as different prices. Because verified third-party quality scores tied specifically to price-ending experiments are not published, the table below applies the ratio’s methodology to a worked illustration holding quality constant — the structure the reader can replicate against Consumer Reports or J.D. Power scores for any real product pair. The point figures are illustrative; the calculation is exact.

Illustrative calculation. Quality held at category median (ratio numerator = 1.0). Perceived price for the 99-ending item applies a ~$15 downward perception on the $1 nominal cut, the low end of the 15–25¢-per-cent estimate scaled to this price level, per Strulov-Shlain (2023). Replace with verified Consumer Reports or J.D. Power scores and live prices to compute for a real product.

The mechanics: the round-ending item sits exactly at fair value, ratio 1.00. The 99-ending item is physically identical but feels like better value — a 1.06 ratio — purely because the perceived price dropped far more than the $1 nominal cut. The buyer captures one real dollar of savings while the brain registers something closer to fifteen. That gap between real and perceived value is the psychological premium the seller extracts, and it is the number that separates this framework from the standard claim that 99-ending pricing exists for accounting reasons. The accounting story explains why prices avoid round numbers; it does not explain a measured demand increase at a higher price.

What most coverage of charm pricing gets wrong

The conventional explanation is that $9.99 looks cheaper than $10 and so people buy more of it. True, but incomplete in a way that misleads. The overlooked finding in the field data is conditional and self-limiting: the effect weakens the more it is used.

Anderson and Simester found the $9-ending lift was strongest on new items, where the customer had the least independent information about value. The price ending functions as a signal — a stand-in for “this is a deal” — and that signal only works when the buyer lacks a better reference point. Strulov-Shlain’s modeling reaches a parallel conclusion: the effect dilutes as 99-endings proliferate across a store. When everything ends in 99, no single 99 signals anything. He calculates that retailers forgo 1 to 4 percent of potential gross profits by responding to left-digit bias with crude rules of thumb rather than optimizing against the actual demand they face.

The practical implication runs opposite to the folklore. The effect is not a universal tax on inattentive shoppers. It is concentrated precisely where information is scarce — new products, unfamiliar categories, items you cannot easily price-check. For a sophisticated buyer, that is a map of exactly where to slow down, and it reframes how to read the broader psychology behind premium pricing.

Methodology

This analysis prioritizes peer-reviewed behavioral economics, consistent with the cluster’s sourcing standard that academic findings carry more weight here than secondary aggregation. The mechanism of the left-digit effect rests on Thomas and Morwitz (2005, Journal of Consumer Research). The quantified magnitude — the central figure of the article — comes from Strulov-Shlain (2023, Review of Economic Studies), chosen as the primary source for the 15-to-25-cent estimate because it is built on real retail scanner data across 3,500 products and 25 chains rather than survey response. Field evidence of demand effects comes from Anderson and Simester (2003, Quantitative Marketing and Economics) and Lacetera, Pope and Sydnor (2012, American Economic Review). The anchoring foundation traces to Tversky and Kahneman (1974, Science) with willingness-to-pay magnitudes from Ariely, Loewenstein and Prelec (2003, Quarterly Journal of Economics).

Where a figure reflects a range across studies or price levels — the 15-to-25-cent perceived-magnitude band, or the $150-to-$210 odometer discontinuity — the range is reported rather than a false point estimate. The Finluxy Price-to-Quality Ratio illustration holds quality at the category median because no third-party quality dataset is tied to price-ending experiments specifically; the calculation method is supplied so a reader can run it against verified Consumer Reports or J.D. Power scores for a real product pair. I reviewed each cited study’s published abstract and results directly rather than relying on secondary summaries of the figures.

What this means for a $150k+ household

At higher absolute price points, the left-digit effect costs more in dollars even though it operates on the same psychology. The relevant decisions are not the $4.99 at the grocery store — the penny saved or perceived there is trivial. They are the $4,999 mattress, the $19,990 kitchen renovation quote, the $1,299-a-month lease, the $49,900 versus $50,000 vehicle. Each is engineered to keep the leftmost digit one tier below the round number, and each is a category where a household at this income level transacts often enough for the cumulative perceived-versus-real gap to matter.

The defensible response is mechanical, not moralistic. When a price sits just below a digit boundary, mentally round it up to the next round number before evaluating — treat the $19,990 quote as $20,000, because that is closer to how the seller priced against your anchor. Apply the heaviest skepticism exactly where the field data says the effect is strongest: new products, unfamiliar categories, and purchases you cannot easily benchmark, which is also where comparison against an independent reference such as private label versus name brand quality data does the most work. The buyer who understands that a 99-ending price is a constructed signal rather than a property of the product is the buyer the effect was never designed to catch — and who can still separate genuine signal from status pricing when a quality differential is real. For a household with the income to absorb a few hundred dollars of digit illusion without noticing, the discipline is worth more than the dollars: it is the habit of pricing the thing, not the tag.