Imagine an aircraft with 180 seats.

The airline sells 185 tickets.

At first, that sounds crazy.

What happens if all 185 passengers show up?

Yet airlines deliberately do this.

Why?

Because airlines know that some passengers who book a flight won't actually show up.

And the economics of aviation makes an empty seat extremely expensive.

An airline has 180 seats.

Whether 180 people fly or 150 people fly, the aircraft still has to:

The flight is going anyway.

So an empty seat represents something the airline can never recover.

Once the aircraft takes off, that seat has effectively expired as inventory.

IATA describes airline seats as a perishable product: once a flight departs, an unsold seat has no future value.

Suppose an airline has a 180-seat aircraft.

Based on historical data, it knows that on a particular route:

Around 5 passengers who book normally don't show up.

If the airline sells exactly 180 tickets, it might regularly depart with something like:

175 passengers.

That's five empty seats.

So the airline might sell:

185 tickets.

If approximately five people don't show up:

185 booked − 5 no-shows = 180 passengers

Perfect.

That's the basic logic behind overbooking.

Airlines don't simply pick a random number.

They use revenue-management systems and historical data to estimate the probability that passengers will actually show up.

The models can consider things such as:

IATA notes that airlines use historical no-show percentages to determine how many additional tickets they can sell.

So an airline might effectively calculate:

180 seats + expected no-shows = number of tickets it can safely sell.

Imagine the airline sells 185 tickets.

185 − 6 = 179 passengers

One seat remains empty.

The airline still loses some potential revenue.

185 − 5 = 180 passengers

Perfect.

185 − 2 = 183 passengers

Now the airline has a problem.

There are only 180 seats.

Three passengers cannot board.

That's when overbooking becomes denied boarding or "bumping."

Airlines generally don't immediately pick three people and throw them off the flight.

In jurisdictions such as the United States, airlines must first ask passengers to voluntarily give up their seats, usually in exchange for compensation and a later flight.

And airlines may increase the offer if they don't get enough volunteers.

Only if there aren't enough volunteers can involuntary denied boarding occur, subject to applicable rules.

Because the airline can make it worth their while.

Imagine the announcement:

Someone who isn't in a hurry might think:

"I'm getting paid to wait a few hours?"

For that passenger, it can be a good deal.

For the airline, paying one passenger compensation may still be cheaper than routinely flying with empty seats.

That's the fundamental economic calculation.

Because the alternative isn't free.

If airlines were forced to sell exactly 180 tickets for a 180-seat aircraft, they would have to accept that some percentage of flights would depart with empty seats because of no-shows.

IATA argues that eliminating overbooking could lead airlines to increase fares or impose stricter ticketing and cancellation conditions to protect themselves against empty seats.

In other words:

Overbooking is partly a way of keeping the aircraft full without requiring every passenger who books to actually show up.

This is where the mathematics becomes fascinating.

If an airline underestimates no-shows:

Empty seats → lost revenue.

If it overestimates no-shows:

Too many passengers → denied boarding → compensation + unhappy customers.

So the airline is constantly trying to find the sweet spot.

It's essentially asking:

That's a revenue-management problem.

Think about an airline's 180-seat aircraft.

If it sells only 180 tickets and five people don't show:

Revenue = 175 passengers

If it sells 185 and five don't show:

Revenue = 180 passengers

The airline has turned seats that would otherwise have flown empty into paying passengers.

But if only two people don't show:

185 − 2 = 183

Now the model has failed.

That's why airlines monitor booking and check-in data right up until departure.

Overbooking isn't really about airlines selling seats that don't exist.

They're selling reservations based on the statistical expectation that some booked passengers won't use their seats.

Most of the time, the prediction works.

Occasionally, it doesn't.

And that's when you hear:

"This flight is oversold. We're looking for volunteers."

This is the key to the entire story.

A hotel room that isn't sold tonight can't be sold tomorrow for tonight.

A restaurant seat at 8 PM can't be sold at 8:05 PM.

And an airplane seat after the aircraft takes off?

Worth exactly ₹0 to the airline.

So airlines use probability to turn expected no-shows into additional revenue.

It's a remarkably sophisticated business model hiding behind something as simple as:

"Would anyone like to take a later flight?"

Because they expect some passengers to cancel or not show up. Overbooking allows airlines to reduce the number of empty seats and use aircraft capacity more efficiently.

The airline first seeks volunteers willing to take another flight in exchange for compensation. If there aren't enough volunteers, passengers may be denied boarding under the rules applicable to that country.

It depends on the jurisdiction, but it is permitted in many major aviation markets. For example, the U.S. Department of Transportation explicitly states that overbooking is not illegal in the United States.

Because a percentage of passengers don't show up. Selling exactly the seat count would mean more flights departing with empty seats, reducing revenue and potentially increasing fares.

So the next time an airline says a flight is "oversold," remember:

They aren't necessarily being careless.

They're running a giant probability calculation.

180 seats.
185 bookings.
A few expected no-shows.
And millions of dollars riding on whether the prediction is right.

That's the business behind the boarding pass.