Loved to Death

The sandstone arch stands against a sky so deeply blue it looks synthetic. Delicate Arch, in southeastern Utah, is the kind of geological accident that makes people believe in something larger than themselves, a freestanding span of Entrada sandstone, 46 feet high, framing the La Sal Mountains in the distance. It took roughly 300 million years to form. On a summer afternoon, several hundred people crowd the bowl beneath it, waiting their turn for a photograph that will look, on their phone screens, almost exactly like everyone else’s.

The trail to Delicate Arch has no shade. It gains 480 feet over 1.5 miles across exposed slickrock. In July 2024, temperatures at the trailhead regularly exceeded 105 degrees. Rangers documented multiple heat-related rescues per week. The biological soil crust that surrounds the trail, a living community of cyanobacteria, mosses, and lichens that took decades to establish and holds the desert together, has been trampled into dust for dozens of feet on either side of the path. Signs ask visitors to stay on the rock. Many do not read the signs.

This is not a story about bad tourists. The people hiking to Delicate Arch are doing exactly what the National Park Service told them to do: come visit your public lands. The problem is that 331.9 million of them took the invitation in a single year,¹ and the system built to manage that relationship is failing at every level simultaneously.

The numbers that define collapse

In 2024, the National Park System recorded 331.9 million recreation visits, an all-time record. That number has been climbing for decades, with particular acceleration after 2010, driven by social media exposure, population growth, and the post-pandemic outdoor recreation surge. The parks were not designed for this volume. Most of the system’s core infrastructure, roads, bridges, water systems, visitor centers, was built during the Mission 66 era, a ten-year construction program that ended in 1966. Much of it was engineered for a fifty-year lifespan. It is now sixty years old, serving three times the visitor volume it was designed to accommodate.

The deferred maintenance backlog across the National Park System stands at approximately $23 billion.² To put that in proportion: the entire annual budget of the National Park Service is roughly $3.3 billion. The backlog is seven times the annual budget. Every year, the gap widens. Infrastructure ages faster than it can be repaired. New damage compounds old damage. The backlog is not a static number waiting to be addressed. It is an accelerating curve.

Congress recognized this problem in 2020, when it passed the Great American Outdoors Act and created the National Parks and Public Lands Legacy Restoration Fund. The GAOA committed $5.3 billion over five years,³ the largest single investment in national park infrastructure in modern history. By the metrics that matter, it was not enough. During the same period that GAOA invested $5.3 billion, the overall backlog grew by $10.3 billion. The funding covered roughly half the rate of new deterioration. It did not touch the existing deficit.

The math is straightforward. When a system’s rate of degradation exceeds its rate of repair by a factor of two, investment does not close the gap. It slows the rate at which the gap opens. This is the structural reality of the national park maintenance crisis, and no amount of celebratory press releases about bipartisan funding achievements changes the arithmetic.

Fewer people managing more people

The infrastructure numbers would be more manageable if the workforce responsible for that infrastructure were stable. It is not.

Since 2010, the National Park Service has lost approximately 40 percent of its workforce.⁴ The decline has been gradual enough to avoid the kind of headlines that accompany sudden mass layoffs, but the cumulative effect has been transformative. In 2025, a further 24 percent reduction in permanent staff was implemented as part of broader federal workforce restructuring.⁵ The parks now operate with staffing levels that bear little relationship to their operational demands.

The workforce reduction is not evenly distributed across functions. It has hit certain categories with particular severity. Law enforcement rangers, the personnel responsible for visitor safety, search and rescue, wildlife protection, and resource crime prevention, numbered approximately 2,300 in 2010. By 2023, that number had fallen below 1,200.⁶ The seasonal law enforcement workforce, which provides critical capacity during peak visitation months, dropped from 825 to 43 over the same period.⁷ Forty-three seasonal law enforcement rangers for the entire National Park System, across 429 units spanning 85 million acres.

The consequences of this reduction are predictable and well-documented. Response times for emergencies increase. Resource crimes, poaching, artifact theft, vandalism of geological and archaeological features, go undetected or uninvestigated. Visitor conflicts escalate without intervention. Search and rescue operations are delayed. In several parks, entire districts go unpatrolled for days at a time during peak season.

Maintenance staff reductions follow a similar pattern. With fewer people available to repair trails, clear drainage, replace aging water systems, and maintain roads, the rate of infrastructure deterioration accelerates. Deferred maintenance is not just a financial abstraction. It is a physical process. A clogged culvert that goes uncleared for one season causes erosion. The erosion undermines a retaining wall. The retaining wall fails and damages a road. What would have been a two-hour maintenance task becomes a $200,000 reconstruction project. Compound interest works on infrastructure decay just as it works on debt.

The workforce crisis also degrades the parks’ capacity for their other core mission: resource stewardship. The National Park Service is the custodian of some of the most significant ecological, geological, and cultural resources on the continent. That custodianship requires monitoring, water quality sampling, wildlife surveys, vegetation assessments, archaeological site documentation. When staffing falls below the threshold needed to maintain monitoring programs, degradation happens invisibly. The data that would reveal the scope of the damage is no longer being collected.

The feedback loop

These three forces, rising visitation, declining infrastructure, and shrinking workforce, do not operate independently. They form a feedback loop that makes each problem worse.

Higher visitation increases wear on infrastructure, accelerating the rate at which it degrades. Degraded infrastructure requires more maintenance, but there are fewer staff to perform it. Understaffed parks cannot manage visitor flows effectively, leading to crowding, resource damage, and safety incidents. Safety incidents and degraded experiences generate political pressure, which is typically directed not at the structural causes, inadequate funding and staffing, but at any measure perceived to restrict public access. The restriction of access is the one tool that would break the feedback loop.

Consider the sequence in practice. A popular trail in a heavily visited park suffers erosion damage from overuse. The trail crew that would normally repair it has been cut in half. The damage worsens over two seasons. A section of the trail becomes hazardous. The park considers a temporary closure for rehabilitation. The closure generates complaints from visitors and outfitters. A congressional representative writes a letter. The park reopens the trail before repairs are complete. The damage accelerates.

This is not a hypothetical. It is the default operating condition at dozens of units across the system.

The environmental toll no one budgets for

The conversation about national park overcrowding tends to focus on the visitor experience, long lines, full parking lots, crowded trails. This framing is incomplete. The environmental consequences of unrestricted visitation are substantial, measurable, and, in some cases, irreversible.

In Yellowstone National Park, researchers have estimated that visitation generates more than one megaton of carbon dioxide annually,⁸ from vehicle emissions within the park, the energy consumption of lodges and visitor facilities, and the transportation impacts of getting 4.5 million people to a remote corner of Wyoming each year. Ninety-seven percent of national park units have significant air pollution levels,⁹ much of it transported from surrounding regions but a meaningful fraction generated by the parks’ own visitor infrastructure.

The ecological impacts extend well beyond air quality. A comprehensive study of wildlife behavior in parks found that 16 of 22 mammal species significantly altered their behavior in the presence of hikers.¹⁰ Animals shifted to nocturnal activity patterns, reduced their use of critical habitat areas near trails, and exhibited chronic stress indicators. These behavioral changes have cascading effects through ecosystems, when large herbivores avoid riparian areas near popular trails, vegetation communities shift, which affects the invertebrate communities that depend on those plants, which affects the fish and amphibian populations that depend on those invertebrates.

In Glacier National Park, the physical evidence of long-term environmental change is starkly visible. In 1850, the park’s namesake landscape contained roughly 150 glaciers. By 2024, approximately 25 named ice bodies remained,¹¹ and many of those no longer met the minimum size threshold to be technically classified as glaciers. Climate change is the primary driver of glacial recession, but the infrastructure of visitation, roads built across fragile alpine terrain, visitor facilities that disrupt hydrological patterns, the heat island effects of parking areas, compounds the problem at the margins.

The biological soil crusts at Arches tell a similar story at a smaller scale. These cryptobiotic communities fix nitrogen, retain moisture, and prevent erosion across the Colorado Plateau. They grow at a rate of roughly one millimeter per year. A single footprint can destroy decades of growth. In areas adjacent to popular trails and viewpoints, the crusts have been effectively eliminated. The exposed soil erodes freely, carrying sediment into drainages and altering the very geology the park was established to protect.

The Arches experiment

Against this backdrop, one park ran a controlled experiment in managing visitation, and it worked.

In 2022, Arches National Park implemented a timed-entry reservation system during its peak season. The system required visitors to book a specific entry window in advance, capping the number of vehicles entering the park during any given period. The goals were straightforward: reduce crowding at popular sites, improve the visitor experience, decrease resource damage from overflow parking and off-trail travel, and bring traffic volumes in line with the park’s infrastructure capacity.

By every available metric, the system succeeded. Visitor surveys showed 89 percent approval,¹² a remarkable number for a restriction on access to a public resource. Traffic flow improved. Emergency incidents related to overcrowding decreased. Resource damage at heavily impacted areas slowed. Rangers reported being able to perform their duties more effectively. The parking areas at Delicate Arch trailhead, which had previously overflowed onto fragile desert surfaces daily during peak months, operated within their designed capacity.

The system was not without legitimate criticisms. Reservation platforms can disadvantage people without reliable internet access or the ability to plan months ahead. International visitors and travelers with flexible itineraries faced logistical challenges. There were equity concerns about a system that effectively required digital literacy and advance planning to access public lands. These are real issues that require thoughtful design solutions, tiered reservation windows, walk-up availability, partnerships with gateway communities to facilitate access.

But the political pressure that ultimately ended the program was not primarily about equity. The reservation system was discontinued for the 2026 season following pressure from local business interests, tourism industry groups, and political representatives who framed any restriction on park access as an infringement on public rights. The framing was effective. The evidence was ignored.

The Arches case is instructive because it demonstrates three things simultaneously. First, that timed-entry systems work, they measurably improve both visitor experience and resource protection. Second, that visitors overwhelmingly support these systems once they experience them. Third, that none of this matters if the political economy favors unrestricted access.

The political economy of public lands

The national parks generate $56.3 billion in total economic output annually,¹³ supporting roughly 415,000 jobs in gateway communities. This economic activity is the parks’ greatest political asset and their greatest structural vulnerability.

The economic argument for parks is their most reliable source of political support. It is also the argument that, taken to its logical conclusion, guarantees their degradation. If the value of parks is measured primarily by visitation and economic output, then any policy that reduces visitation, however beneficial to the parks themselves, can be framed as economically harmful. This creates a structural bias against the management tools that the parks most need.

Gateway communities, tourism operators, and hospitality businesses have rational economic incentives to maximize visitor throughput. Their income depends on volume. A timed-entry system that reduces peak visitation by 15 percent represents a 15 percent reduction in potential revenue for the businesses that serve those visitors. The long-term argument, that overcrowding degrades the very resource that generates the economic activity, is correct, but it operates on a timeline that does not align with quarterly revenue targets.

Congressional oversight of the parks reflects this dynamic. The most common form of congressional engagement with the National Park Service is a letter from a member of Congress complaining that constituents were unable to access a park, could not find parking, or experienced long wait times. These letters generate bureaucratic responses, which generate caution, which generates institutional reluctance to implement access management tools. The parks have learned, through decades of experience, that restricting access carries political costs that expanding access does not, regardless of what the resource data says.

The Great Smoky Mountains National Park illustrates this dynamic at its extreme. Great Smoky is the most visited park in the system, recording 12.19 million recreation visits in 2024.¹⁴ It is also one of the few major parks that does not charge an entrance fee, a consequence of the terms under which the land was donated and the road system was built. The park has explored various congestion management tools, but each proposal faces the additional political freight of being perceived as imposing costs on visitors to a historically free park.

The result is a park that absorbs more human traffic than any other unit in the system, with funding that does not reflect that volume, and management tools that are politically constrained. It is, in structural terms, a commons that is being consumed faster than it can regenerate.

What the science says about carrying capacity

The National Park Service’s Organic Act of 1916 established a dual mandate: to conserve park resources and to provide for public enjoyment. For most of the parks’ history, these two goals were treated as complementary. The more people who enjoyed the parks, the broader the constituency for conservation. This assumption held when visitation was measured in the tens of millions. It does not hold at 331.9 million.

Ecological carrying capacity is not a simple number. It varies by ecosystem, by season, by the type and intensity of use, and by the resilience of specific resources. But the concept is not infinitely elastic. Every ecosystem has thresholds beyond which use causes damage that exceeds the system’s capacity for recovery. Dozens of parks have documented that current visitation exceeds these thresholds for specific resources, trails, water systems, wildlife habitat, air quality, soundscapes.

The challenge is that carrying capacity is defined by science, but access is governed by politics. The two systems operate on different logics. Science deals in thresholds, tolerances, and empirical evidence of degradation. Politics deals in constituencies, economic interests, and the rhetorical framing of public rights. When a senator says that the parks belong to the people and should not be rationed, the statement is politically effective regardless of its relationship to ecological reality.

The tension is not new. What is new is the scale. At 331.9 million visits per year, with visitation projected to continue climbing as population grows and international tourism recovers from pandemic disruptions, the margin for political accommodation has narrowed. The parks cannot simultaneously provide unrestricted access and preserve the resources that make them worth visiting. This is not an ideological position. It is a physical constraint.

The model that exists

The irony of the national park crisis is that the solution has already been demonstrated. Timed-entry reservation systems, implemented at Arches, Rocky Mountain, Glacier, Haleakala, and several other parks over the past five years, have consistently produced the same results: improved visitor experience, reduced resource damage, high public approval, and more effective park operations.

These systems are not unprecedented in the management of public resources. Wilderness areas across the National Forest System have used permit systems for decades. River corridors on the Colorado, the Salmon, and dozens of other waterways manage access through lottery and reservation systems that are widely accepted. Campgrounds throughout the federal land system operate on a reservation basis. The principle that public resources can be managed through access controls without ceasing to be public is well-established in American land management.

What makes the parks different is visibility. A permit system for a wilderness area that receives 10,000 visitors a year does not generate congressional letters. A reservation system for a park that receives 4 million visitors a year generates headlines. The political cost of access management scales with the popularity of the resource, which means that the parks most in need of management are the parks where management is hardest to implement.

The system-wide solution would involve several interlocking elements. Universal timed-entry reservation systems during peak seasons at high-visitation parks. Dynamic pricing that distributes demand across seasons and days of the week. Investment in gateway community infrastructure to distribute visitor impact. Restoration of the workforce to levels adequate for management, maintenance, and law enforcement. Sustained maintenance funding at a scale that actually reduces, rather than merely slows the growth of, the backlog.

Each of these elements has been tested. Each has evidence supporting its effectiveness. None faces primarily technical barriers to implementation. The barriers are political, and the political incentives currently favor inaction.

What compounding looks like

Systems under stress do not fail linearly. They fail through compounding. Each unaddressed problem intensifies the others, and the interactions between failures create new failure modes that were not present when each problem existed in isolation.

The national park system is in the early stages of compounding failure. The most immediate manifestation is the visitor experience itself. Parks that once offered solitude, immersion in landscape, and the particular quality of attention that comes from being in a wild place increasingly offer something closer to a theme park experience, lines, crowds, noise, and the constant presence of other people’s recording devices. This is not a trivial loss. The experiential qualities that distinguish national parks from other forms of recreation are precisely the qualities that overcrowding eliminates.

But the deeper compounding is ecological. When monitoring programs are cut, the data that would document environmental decline stops being collected. When maintenance staff are reduced, the infrastructure that channels and contains human impact fails, spreading that impact across broader areas. When law enforcement is gutted, the regulatory framework that protects resources exists only on paper. When political pressure prevents access management, the underlying cause of all these downstream effects, too many people in too little space with too few resources for management, continues unabated.

Glacier’s ice is not coming back. The biological soil crusts at Arches will not regenerate in a human lifetime. The behavioral disruption of wildlife populations has demographic consequences that compound over generations. The infrastructure that was built sixty years ago and has been deteriorating for twenty will not become easier or cheaper to repair next year.

The national parks were established on a premise: that some landscapes are valuable enough to warrant permanent protection from the pressures of commerce and development. That premise has not changed. What has changed is the nature of the threat. The parks are not being drilled, mined, logged, or dammed. They are being loved, and the system designed to manage that love is being systematically starved of the capacity to do so.

The tools to address this exist. They have been tested. They work. The question is whether the political system is capable of implementing solutions that require telling a portion of the public, in any given moment, that they will need to wait their turn.

The parks were built for permanence. They are being managed for the next election cycle. The distance between those two timeframes is where the damage accumulates.