The Reservation That Makes the Crowd

The system built to ration demand can also manufacture scarcity, commitment, and congestion.

Technology, when overused, rings a small funeral bell for freedom and poetry in ordinary life.

Yosemite used to be a place we visited with very little ceremony. We decided to go, bought a ticket at the gate, and went in. We have returned almost every summer for years. Then came reservations, lotteries, release dates, time windows, and the familiar modern ritual of sitting in front of a browser at precisely the hour when thousands of other people are sitting in front of theirs.

The park remained magnificent. Access, however, had acquired the administrative charm of a tax return and the punctuality requirements of minor surgery.

During the reservation years, Yosemite also seemed much more crowded to me. The usual explanation is straightforward: the park introduced reservations because too many people wanted to visit. The system was managing the crowd that already existed.

That explanation is perfectly sensible. It also fails to match my experience.

This year, Yosemite does not require a vehicle reservation. The National Park Service dropped the requirement after reviewing 2025 traffic and parking data and concluding that most weekdays had available parking, stable traffic flow, and visitation within operational capacity. We had a sudden idea, booked a hotel, by which I mean a tent with a billing department, and went.

It was lovely. People were around, of course, but the place did not feel besieged. Early one morning, we walked across the meadow. The air was cool, the grass was quiet, and nobody appeared to be fighting an app.

My immediate conclusion was unscientific and satisfying: perhaps the reservation had been causing some of the crowd.

The feeling extends beyond Yosemite. Once an attraction introduces reservations, one of two things seems to happen. Either the reservation is impossible to obtain, or you obtain one and arrive to find half the country standing in front of you. Occasionally the system achieves both. You fail to get a reservation, reach the site anyway, and discover that the supposedly scarce area is nearly empty.

That happened to us at Angels Landing in Zion. We could not get a permit. We hiked to the point where only permit holders could continue and spoke with the ranger there. Beyond the checkpoint, very few people seemed to be on the trail.

This was a moment, not a dataset. Zion says its permit program has reduced crowding and congestion on the final half-mile of the trail, which is a reasonable objective for a route where a bad passing maneuver has unusually serious consequences. The permit may be doing exactly what it was designed to do.

Still, there is something exquisitely modern about being denied entry to an empty place by a full database.

So is my rule of thumb, “once reservations begin, good luck getting in,” merely nostalgia wearing hiking shoes? Possibly. Yet several branches of economics suggest that reservation systems can do more than distribute existing demand. They can alter the demand itself.

The most obvious mechanism is scarcity.

A reservation calendar takes a physical limit that was previously vague and turns it into visible information. Twelve spaces left. Two dates available. Summer weekends sold out. Next release at 7 a.m. Pacific time. The resource may always have been limited, but the website has now placed the limitation in a glass case and illuminated it.

A 2022 meta-analysis examined 416 effect sizes from 131 studies on scarcity. Scarcity cues generally increased desirability and purchase intention. Particularly relevant here, supply-based scarcity had a large effect on experiences.

The practical result is familiar. “Perhaps we should visit someday” becomes “There are only six weekends left.” The destination has acquired urgency without adding a single waterfall.

Tourism researchers have found that the source of scarcity matters too. “Only three rooms left because everyone is booking” sends a different psychological signal from “only three rooms exist.” In experiments on online booking, demand-framed scarcity could raise purchase intention by signaling popularity, with the effect depending on the consumer and perceived risk.

A reservation system therefore communicates two things at once. It tells us what is available, and it tells us what everyone else appears to want.

That second message matters because booking status can become social proof. Isabel Kaluza, Guido Voigt, and Friederike Paetz ran experiments in which consumers chose between service providers whose calendars showed different levels of occupancy. An empty calendar could look suspicious. A partly booked calendar suggested quality. An almost full calendar eventually became unattractive because flexibility disappeared. The preference followed an inverted U shape: busy was good, until busy became unbearable.

It is the empty-restaurant problem transferred to a screen. We assume that the diners already inside know something we do not. A modest queue is a recommendation. An empty room raises questions. A queue around the block eventually reminds us that sandwiches have been invented elsewhere.

An online reservation page performs this signaling before we leave home. Every unavailable date tells the next visitor that somebody else considered the experience worth having. Popularity becomes evidence of quality, which creates more popularity. The calendar is doing marketing while dressed as administration.

There is even research suggesting that service providers can benefit from controlling how much availability consumers see. When abundant open appointments are interpreted as evidence of low quality, releasing fewer visible slots can improve demand. At that point the distinction between capacity management and theater becomes rather delicate.

Wanting the reservation, of course, is still different from needing to go. That distinction begins to disappear once you have one.

A person without a reservation has a preference. A person with a reservation has a plan. The latter has selected a date, coordinated with companions, arranged transportation, perhaps booked a room, and possibly explained to a manager why Friday afternoon has suddenly become a matter of geological importance.

A randomized experiment published in The Economic Journal gives this mechanism a rather stark demonstration. Researchers offered HIV-testing appointments to high-risk men in Malawi. Simply giving people appointments more than doubled testing rates, with much of the effect concentrated among people who appeared to value help overcoming procrastination and self-control problems. The authors describe appointments as commitment devices.

A medical test and Yosemite are obviously different activities. The behavioral point travels rather well. An intention that has acquired a date becomes harder to abandon.

Before reservations, a family might wake up, look at the weather, reconsider a four-hour drive, and stay home. After reservations, the same family has a scarce Saturday entry, two hotel nights, coordinated schedules, and a small emotional investment in having defeated the website. Rain now looks atmospheric.

Reservations also lower uncertainty. The possibility of driving several hours and failing to get in is itself a cost. Guarantee admission and that cost falls. Some people who would never gamble an entire weekend on uncertain access will happily travel once the uncertainty has been removed.

So a system designed to restrict capacity can simultaneously make the trip more attractive.

Then we get the opposite problem: people reserve and do not come.

When the cost of reserving is low, a permit has option value. I may not know whether I want to go on August 12, but I know I would rather possess the option than discover on August 10 that someone else has it. So I book now and decide later.

Operations researchers have long studied this problem in restaurants and other queues. Reservations organize demand, yet no-shows waste capacity when those slots cannot be reassigned. The result is an awkward balancing act involving deposits, walk-ins, overbooking, and how much capacity should be reserved in advance.

This does not prove that no-shows explain what I saw at Angels Landing. It does explain how a place can be “sold out” while appearing surprisingly empty. The database contains intentions. The trail contains people.

Free reservations can become objects of competition in their own right. An American Economic Review paper documented black markets in free government appointment systems, where scalpers captured slots and resold them. Its authors showed theoretically and experimentally that collecting applications in batches and allocating appointments randomly could eliminate profitable scalping under the conditions they studied.

National park visitors are generally not passport-appointment scalpers. The interesting point is simpler. Once a reservation becomes scarce and valuable, people begin competing to possess the reservation itself. Whether that reservation eventually turns into the intended activity is a separate question.

Reservations also synchronize us.

Without reservations, arrivals are messy. Some people leave early, some late, some change their minds after breakfast, and some take a wrong turn because their phone has strong opinions about mountain roads. Reservation systems introduce common release dates, entry windows, deadlines, and incentives. Thousands of independent people suddenly share the same clock.

Queueing economists call one version of this “strategic arrivals.” People choose when to arrive partly by anticipating delays and the behavior of everybody else. Individually sensible decisions can produce clustered equilibrium arrival patterns. Perfectly reasonable individual behavior generates an unreasonable crowd. Everyday life usually calls it Saturday morning.

This distinction matters because total attendance and experienced crowding are not the same quantity. A reservation system can reduce disastrous peaks while also eliminating quiet periods. It can produce fewer impossible days and more reliably full ones. Annual visitation can remain similar while visitors increasingly occupy the same parking lots, viewpoints, shuttle stops, and trail sections at the same time.

As for Yosemite, it supplied a rather neat example. In 2022, reservations were required between 6 a.m. and 4 p.m. Visitors without one could simply enter before or after that window. They did. The National Park Service later described “shifting use patterns” under the 2022 system and moved the morning boundary to 5 a.m. in 2024. Its own planning documents note that more vehicles were entering before peak hours began. The reservation had regulated one crowd and, quite inadvertently, scheduled another.

There is no single paper showing that reservations generally make public attractions more crowded. The literature is modular. Scarcity increases desire. Visible booking levels transmit social proof. Appointments improve follow-through. Cheap reservations create options and no-shows. Common windows synchronize arrivals.

Put the pieces together and my Yosemite impression stops looking entirely eccentric. A mechanism introduced to ration demand can also stimulate demand, convert casual intentions into actual visits, and concentrate those visits into predictable moments.

And none of this is really confined to parks.

“Few spots left” is among the most durable sentences in modern commerce. It sells restaurant tables, concert tickets, online courses, hotel rooms, sneakers, and things whose manufacturers have heroically discovered that a factory is capable of stopping after exactly 500 units.

A national park at least has a real carrying capacity. A “limited edition” has whatever carrying capacity someone in a marketing meeting decided to type into a spreadsheet. In one case scarcity is discovered. In the other, it is authored.

Let me derail for a moment to a recent movie: Christopher Nolan’s The Odyssey.

The film was shot entirely with IMAX film cameras, and IMAX 70mm exhibition is genuinely scarce. At the moment, there are only about forty theaters in the world showing it that way, 24 in the United States, eight in California, and two in the San Francisco Bay Area.

The Bay Area has roughly 7.7 million residents. Now imagine those two screens operating with almost comic efficiency. The Odyssey runs for 172 minutes. If each theater somehow managed eight screenings a day, every day, with barely enough time between them to sweep up the popcorn and apologize to the projector, that would produce 112 screenings a week. Assume, generously and for easy arithmetic, 350 seats per screening. We still get fewer than 40,000 seats a week.

And that is the fantasy capacity. Actual theaters need cleaning, maintenance, projection checks, staff, and occasionally sleep. The Metreon has in fact been running screenings into the small hours because demand is so high.

That is the level of scarcity we are talking about.

Now consider what exactly is scarce.

For a director, narrative is the soul of the work. Technology is the channel through which it reaches us. A better channel can reveal more detail, greater scale, richer sound, and a different physical experience. That can make one presentation superior without making every lesser presentation artistically invalid.

Otherwise, why do we still watch Interstellar at home? Why does anyone rent 12 Angry Men?

The distribution itself rather gives the game away. IMAX describes The Odyssey as opening “everywhere,” despite the film having been shot entirely with IMAX cameras. On opening weekend, IMAX accounted for 20 percent of the film’s global box office. That is an extraordinary number for one premium format. It also means that roughly four-fifths of the money was spent by people watching the film somewhere else.

If an ordinary cinema were incapable of carrying the artistic work, releasing the movie to most of the world that way would amount to authorizing a defective edition at industrial scale. That seems an unnecessarily eccentric theory of authorship.

IMAX 70mm can plainly be the best version. It does not follow that it is the only legitimate one.

Marketing has little incentive to preserve that distinction. “The picture is larger and the sound more immersive” describes a premium product. Perfectly respectable, perhaps even compelling. “This is how the director intended you to see it” does something more useful. It quietly suggests that without the scarce ticket, you have not quite seen the movie at all.

The scarcity of the screens is real. The necessity attached to them is a separate product.

This is the same behavioral lever in rather more elegant clothing. Which is why, despite having written all of the above, I still have not seen The Odyssey.

I know perfectly well what the scarcity is doing to me. I am still waiting for an IMAX 70mm ticket. Understanding behavioral economics, regrettably, does not confer immunity from it.

The operator sees capacity management. The visitor sees a mountain that has somehow acquired a login screen. The cinema sees premium presentation. The customer sees six remaining seats and begins wondering whether Tuesday at 10:45 p.m. is actually quite convenient.

As consumers, understanding the mechanism offers disappointingly little tactical advantage. A booking page does not tell us whether scarcity is physical, administrative, strategic, temporary, or partly created by people reserving something they will never use. We see the red calendar. We act.

As producers, the conclusion is rather more cheerful. Scarcity is powerful. Commitment is powerful. Social proof is powerful. A reservation can regulate demand and stimulate it at the same time. This is useful information if you run a park, a restaurant, a cinema, or almost anything else with a calendar and customers.

Which is why this is, first, a behavioral economics phenomenon, and then a rather different lesson depending on which side of the booking page you happen to occupy.

Yosemite happened to remove its vehicle reservation this year. We booked our tent, crossed the meadow in the morning, and found freedom and poetry still wandering around without an appointment.

Good luck to us all.

Sources

Barton, Zlatevska, and Oppewal, “Scarcity Tactics in Marketing: A Meta-analysis of Product Scarcity Effects on Consumer Purchase Intentions.”

Huang, Liu, Kandampully, and Bujisic, “Consumer Responses to Scarcity Appeals in Online Booking.”

Kaluza, Voigt, and Paetz, “Empirical Studies on the Impact of Booking Status on Customers’ Choice Behavior in Online Appointment Systems.”

Kaluza, Voigt, Haase, and Dietze, “Control of Online-Appointment Systems When the Booking Status Signals Quality of Service.”

Sterck, “Healthcare Appointments as Commitment Devices.”

Hakimov, Heller, Kübler, and Kurino, “How to Avoid Black Markets for Appointments with Online Booking Systems.”

Oh and Su, “Reservation Policies in Queues: Advance Deposits, Spot Prices, and Capacity Allocation.”

Honnappa and Jain, “Strategic Arrivals into Queueing Networks: The Network Concert Queueing Game.”