Cross Price Elasticity Calculator

What cross price elasticity measures

Cross price elasticity of demand answers a narrow question: when the price of one product moves, how much does the quantity sold of a different product move? Write the two goods as X and Y, and the formula is the percent change in quantity of X divided by the percent change in the price of Y.

The sign carries the meaning, and it is the part people skip. When the price of Y rises and buyers switch toward X, quantity of X rises and the elasticity is positive: the goods are substitutes. When the price of Y rises and buyers use less X as a result, the elasticity is negative: the goods are complements, because they are consumed together. A number near zero says the two products barely notice each other.

Cross price elasticity is not the same measure as own price elasticity, which compares a product's quantity with its own price and is almost always negative. A shop can report that coffee has an own price elasticity of -0.67 and that tea has a cross price elasticity of +0.80 with respect to coffee. Both numbers describe demand, and they answer different questions.

One property matters for interpretation: the two directions are usually not equal. The elasticity of tea with respect to coffee is not the same number as the elasticity of coffee with respect to tea, because the two markets differ in size, in spend, and in how easy it is to switch. Quote the direction you measured, not just the pair.

Worked example: coffee and tea at one cafe

A cafe sells 500 cups of tea a week at $3.00 while the espresso bar charges $4.00 for a latte. The latte price goes to $4.60, and tea sales climb to 560 cups a week. Treat the latte as Y and tea as X.

Quantity of tea moves from 500 to 560, a change of 60 cups. Price of the latte moves from $4.00 to $4.60, a change of $0.60. Using the starting values as the base, 60 / 500 = 0.12 and 0.60 / 4.00 = 0.15. Divide the first by the second: 0.12 / 0.15 = 0.80. Tea and lattes are substitutes, and a 1 percent latte price rise is tied to a 0.8 percent tea sales rise.

The two formulas land within a hundredth of each other here because the price move is small, about 15 percent. That is the usual result. The next example shows what happens when the move gets bigger.

Worked example: a printer and its ink

An office store cuts the price of a home printer from $120 to $102, down 15 percent. Ink cartridge sales at the same store rise from 900 to 990 units a month, up 10 percent. Printer is Y, ink is X.

Starting base: 90 / 900 = 0.10 for ink, and 18 / 120 = 0.15 for the printer. The elasticity is 0.10 / -0.15 = -0.67. The negative sign says the two products are complements: cheaper printers put more ink machines on desks, and each one pulls cartridges.

Midpoint base: the average ink quantity is 945, so 90 / 945 = 0.0952. The average printer price is $111, so 18 / 111 = 0.1622. Divide: 0.0952 / -0.1622 = -0.59. The gap between -0.67 and -0.59 is larger than in the coffee case because the price fell 15 percent relative to a smaller price level, which is exactly the situation the midpoint formula was built for.

Both numbers pass the same test: negative, so complements, and below 1 in absolute value, so the link is real but not dominant. A store that assumed ink sales track printer sales one for one would overstock.

Reading the number: substitutes, complements and noise

The bands below are a working convention used in pricing and in teaching, not a rule written into any statute. They are a starting point for a conversation, and the text after the table explains why small values deserve suspicion rather than excitement.

Published estimates tend to be smaller than people expect. The U.S. Department of Agriculture's Commodity and Food Elasticities database reports cross price elasticities for bread that run from -0.0223 to 0.0739, with a mean of 0.0254. Note the date on that source: the database stopped being updated and its documentation carries a last update of February 2006, so it is a reference point for method and magnitude rather than a current market reading.

Values near zero are not a failure of the method. Most pairs of goods in a grocery basket are close to unrelated, and a measured 0.03 is a realistic answer for two items that share a shelf but not a use.

Starting value or midpoint: which formula to use

There are two common ways to turn prices and quantities into percent changes, and they give different answers whenever the change is large.

The starting value method divides the change by the value you began with. It is simple, it matches how people read a price tag, and it is fine for moves under roughly 10 percent. The midpoint method divides the change by the average of the before and after values. It gives the same answer whichever direction you travel, which matters when you compare two studies or two years of data.

The second row is the warning. Doubling the price change from 15 to 50 percent does not scale the starting value answer in a straight line, because 0.20 / 0.50 = 0.40 uses a quantity base of 500 while the midpoint uses 550. Report the method with the number, and never mix a starting value elasticity from one study with a midpoint elasticity from another.

There is a third version, used in economics papers: point elasticity, written as the derivative of quantity with respect to price, multiplied by price over quantity. For a straight line demand curve it changes at every point, which is a reminder that an elasticity is a local measurement, not a fixed property of a product.

Using elasticity to set a price

Flip the formula and it becomes a planning tool. If you want quantity of X to rise by some percentage and you know the cross elasticity, the price of Y has to move by the target percent change divided by the elasticity.

Back at the cafe: tea is at 500 cups and the owner wants 600, a 20 percent gain. With an elasticity of 0.80, the latte price has to rise by 0.20 / 0.80 = 0.25, or 25 percent, which takes $4.00 to $5.00. That is a big jump for a coffee, and the plan carries an obvious cost: latte sales fall when the latte gets more expensive. At an own price elasticity of -0.67, a 25 percent rise cuts latte quantity by 16.75 percent, from 400 drinks a week to 333.

Count the money both ways. Tea revenue at 600 cups and $3.00 is $1,800 a week, up $300 from 500 cups. Latte revenue at $5.00 and 333 drinks is $1,665, up $65 from 400 drinks at $4.00. The combined gain is $365 a week on paper, before the milk, cups and labor that 100 extra teas require.

That arithmetic is why cross elasticity is a diagnostic, not a pricing strategy on its own. It tells you which lever moves which product and roughly how far. It does not tell you whether the trade is profitable, because that depends on margins, capacity, and whether regulars come back after the latte hits five dollars.

Where the number breaks down

An elasticity is a ratio of two changes measured while everything else is held still. In a real shop nothing else is held still, and these are the usual ways the estimate goes wrong.

Other things moved too. The week the latte price rose might also have been cold, or a competitor closed, or a promotion ran. A cross elasticity computed from two weeks of data absorbs all of that into one number.

Cause runs both ways. If tea sales were already climbing, the cafe may have raised the latte price in response to higher traffic. The correlation is real, the direction is not.

The level of aggregation changes the answer. Tea against coffee at category level is one number. A specific chai latte against a specific espresso drink is another, usually larger, because narrow products are closer substitutes.

The base is tiny. Going from 5 units to 6 units is a 20 percent change on one extra sale. Elasticity divides that noise by another small number and can produce a spectacular result that means nothing.

The time horizon matters. Buyers adjust slowly. A one month window understates substitution because habits, menus and shelf space take longer to change, so short run elasticities usually sit closer to zero than long run ones.

The relationship itself moves. Streaming and cable were substitutes a decade ago and now sit in a different market. An elasticity is a snapshot with a date, and a five year old number is a historical document.

Why regulators and category buyers both use it

Two different groups rely on this measure, and they use it for opposite purposes.

Competition agencies use it to draw market boundaries. The Horizontal Merger Guidelines published by the Federal Trade Commission and the Department of Justice in 2010 describe a hypothetical monopolist test: would a single seller of a candidate group of products find it profitable to raise price by a small but significant and non-transitory amount, commonly 5 percent? If buyers would switch away in volume to a product outside the group, that product belongs inside the market. The switching is measured with cross elasticities and diversion ratios, so a merger case can turn on whether two products are close enough substitutes.

Retail buyers use it closer to home: which items cannibalize a promotion, how much of a price cut on one item gets recovered on the attachment, and whether a new SKU will grow the category or just move sales around inside it.

For a small business the practical version is a two week test. Change one price, hold everything else, record unit sales of the related product, and compute the number both ways. Even a rough estimate with a stated method beats a guess, and it gives you a baseline to update later. If you are checking a percentage change by hand, the percentage calculator handles the arithmetic, and unit price tells you whether the margin moved with the volume.