Why Can't We Build Anymore?
The West Doesn't Seem Able to Build Megaprojects Anymore
While India erects the world's highest railway bridge through the Himalayas and China lays 48,000 kilometers of high-speed rail, the West is getting buried alive in its own process. A forensic examination of how the most powerful economies on Earth forgot how to build things.
Bottom Line Up Front: The United States, United Kingdom, and most of Western Europe have developed a structural inability to build major infrastructure on time, on budget, or sometimes at all. California's High-Speed Rail — through flat Central Valley farmland — now costs up to $128 billion for 500 miles with no completion date in sight. Britain's HS2 has consumed £40.5 billion ($52B) by April 2025, its northern leg has been cancelled, and its chief executive wrote to the Transport Secretary in March 2025 calling the organization's performance a fundamental failure. New York City's Second Avenue Subway costs $2.6 billion per mile — 30 times the cost of equivalent construction in Spain or South Korea. The causes are not a mystery: they are documented, peer-reviewed, and bipartisan. They include environmental review weaponized as obstruction, procurement designed to transfer risk rather than build things, hollowed-out public-sector technical capacity replaced by consultants who bill by the hour, gold-plated overspecification, and a political culture that treats opposition to infrastructure as a legitimate permanent veto. The nations still building — China, India, Spain, South Korea, Turkey — share one trait the West has largely lost: the institutional will to decide that a project is necessary and then clear the path for it.
In the summer of 1869, a golden spike was driven into a railroad tie at Promontory Summit, Utah, and the United States was permanently transformed. The transcontinental railroad — built in six years, through the Sierra Nevada and the Rockies, without power tools, without computers, without modern explosives until Alfred Nobel's dynamite arrived mid-project — knitted the continent together and made the American century possible. Twenty-two years later, the interstate highway system moved from dream to coast-to-coast reality in under fifteen years. Hoover Dam went from authorization to power generation in five years and under budget. The Pentagon was built in sixteen months. The Empire State Building in thirteen.
Something has gone badly wrong since then. The same civilization that blasted tunnels through solid granite with hand drills and black powder, that strung a railroad over the Sierra Nevada in winter, that raised a 726-foot concrete dam in a desert canyon — this civilization now struggles to build a train line through the Central Valley of California, the flattest, most geologically benign real estate on the continent, in under four decades at any price it can name with a straight face. This is not a funding problem. It is not a technology problem. It is a civilizational failure of institutional capability, and it has been documented with increasing precision by researchers who find the same pathologies recurring across every project, every jurisdiction, every political administration that tries to build something large in the English-speaking West.
The numbers make the case more starkly than any editorial. India built the world's highest railway bridge — 359 meters above a Himalayan gorge, through Seismic Zone V, in a region with no roads, requiring 26 kilometers of access road construction before a single arch segment could be lifted — for approximately $180 million. The entire 272 km Udhampur–Srinagar–Baramulla Rail Link, including 36 tunnels and 943 bridges through the Himalayas, cost roughly $5.3 billion: about $19 million per kilometer. China has built 48,000 kilometers of high-speed railway since 2008 at an average cost of $17–21 million per kilometer. Spain and France build high-speed rail for $25–39 million per kilometer. California is building high-speed rail through flat farmland for a projected $100–160 million per kilometer — and the project began construction in 2015 with no revenue train yet running, no completion date credibly established, and a cost estimate that has approximately tripled in real terms since the 2008 ballot measure that authorized it.
California's High-Speed Rail: A Case Study in Institutional Decay
California voters approved Proposition 1A in November 2008 by a margin of 52.7 to 47.3 percent, authorizing $9.95 billion in state bonds for a high-speed rail system connecting San Francisco to Los Angeles in under three hours. The ballot measure promised fares of around $55 and a completion date of 2020. It was, at the time, a genuinely exciting proposition: California as the American laboratory for the European and Japanese rail model, a state with the economic heft and political will to demonstrate that the United States could build world-class infrastructure again.
What followed is one of the most extensively documented infrastructure failures in American history, tracked in biennial reports by the California High-Speed Rail Authority itself, reviewed quarterly by a legislative Inspector General, and subjected to congressional scrutiny from both parties. The project's cost estimate has escalated through at least eight major revisions: from $33 billion in 2008 to a current range of $89–128 billion for the full San Francisco–Los Angeles Phase 1 system, with the Authority's August 2025 Supplemental Project Update still unable to provide a funded completion pathway for the full route. By August 2025, $13.8 billion had been spent, concentrated almost entirely on the Initial Operating Segment — 119 miles of new construction in the Central Valley between Merced and Bakersfield, the easiest terrain on the entire route. Revenue service on this segment alone is now projected for 2032. San Francisco to Los Angeles, if it happens at all, is a 2040s proposition at best.
The project also lost $4 billion in federal grant funding during the Trump administration, which California is contesting through litigation. A 13.5-mile Pacheco Pass tunnel required to connect the Central Valley to the Bay Area — potentially the longest passenger rail tunnel in U.S. history — has not yet been contracted. The mountain crossings at both ends of the route, the Tehachapi Pass to the south and the Diablo Range to the north, remain entirely unfinanced. The Authority's own Inspector General, reviewing the August 2025 supplemental report, found that the document still does not provide required information about how the Authority will finance ongoing construction to keep the Merced–Bakersfield segment on schedule for 2032.
2008: Ballot measure promises full SF–LA system for ~$33 billion by 2020.
2011: Cost estimate rises to $43 billion; completion date revised to 2033.
2012: Estimate revised to $68 billion; first major scope reductions.
2018: Authority introduces cost ranges for first time: $63–98 billion.
2022: Estimate: $105–128 billion. Northern leg scope reduced.
2025: Full Phase 1 estimate: $89–128 billion. IOS (119 miles, flat Central Valley) alone: $36.75 billion. No funded pathway for full route. Revenue service on IOS projected 2032.
Total spent by August 2025: $13.8 billion. Miles of track available for revenue service: zero.
HS2: "An Appalling Mess"
Britain's HS2 — High Speed 2, a railway intended to connect London with Birmingham, Manchester, and Leeds — was announced in 2009 by the Labour government with a cost estimate of £37.5 billion and a promise of transformative national connectivity. By any reasonable measure, it has become the most comprehensively documented infrastructure governance failure in British history, scrutinized by the National Audit Office, the House of Commons Public Accounts Committee, and a succession of independent reviews whose conclusions have progressively darkened from "concerned" to "appalling."
The northern leg — Birmingham to Manchester and Birmingham to Leeds — was cancelled by Prime Minister Rishi Sunak in October 2023, in a conference speech delivered in Manchester, eliminating the routes that carried the project's strongest economic justification. The estimated cost of completing just Phase 1 — London to Birmingham, 135 miles — has since risen to £81 billion at 2019 prices, which with construction industry inflation of at least 25 percent since 2019 implies a current-price cost approaching or exceeding £100 billion. By April 2025, £40.5 billion had been spent. HS2's new chief executive, Mark Wild, appointed in December 2024, wrote to Transport Secretary Heidi Alexander on 31 March 2025 stating plainly that "the organisation has failed in its mission to control costs and deliver to schedule." The Transport Secretary's response to Parliament the same day called the situation "an appalling mess" and acknowledged that the 2033 completion target was impossible.
The House of Commons Public Accounts Committee, in January 2024, found that HS2 "now offers very poor value for money to the taxpayer" and that the Department for Transport "does not yet know what it expects the final benefits of the programme to be." The Eno Center for Transportation's Transit Costs Project database lists HS2 as the most expensive high-speed rail project per kilometer in its entire global dataset. The Stewart Review, published June 2025, concluded that "the top-down vision of building a railway that would be the best and fastest has been a major factor in undermining attempts to introduce a culture of cost control" and that the project had been "subject to evolving political aims, which pushed forward on the schedule before there was sufficient design maturity."
Among HS2's more memorable specific expenditures: an elaborate tunnel system designed to protect a bat colony, which the Department for Transport acknowledged had been poorly managed at disproportionate cost — and which the Public Accounts Committee used as a case study for how environmental obligations, however legitimate in principle, had become instruments of budget destruction in practice. The bats cost more to protect than many complete infrastructure projects in countries that are actually building things.
"The top-down vision of building a railway that would be the best and fastest has been a major factor in undermining attempts to introduce a culture of cost control."
Stewart Review, June 2025 — Independent Review of HS2 Corporate GovernanceNew York City: The Extreme Case
If California and Britain represent the middle of the Western infrastructure pathology, New York City represents its terminal expression. The Second Avenue Subway Phase 1 — three stations, 1.8 miles of tunnel on Manhattan's Upper East Side — was opened in January 2017 after decades of delay and at a cost of approximately $4.5 billion, or $2.5 billion per mile. Phase 2, extending the line by 2.4 kilometers northward to 125th Street, is currently projected at $7.7 billion — roughly $3.2 billion per kilometer. That figure is more than triple the cost of equivalent construction in the most expensive non-American country in the NYU Transit Costs Project database, and approximately 30 times the cost of building a comparable subway in Spain, South Korea, or Finland.
The raw materials are not more expensive in New York. Steel, concrete, and copper cost roughly the same on global commodity markets whether the project is in Madrid or Manhattan. What costs more in New York is everything else: station sizes that are functionally three times larger than they need to be, with elaborate mezzanines whose primary value is aesthetic; a consultant-heavy project management structure in which public agencies have outsourced their own technical capacity to firms that bill by the hour and whose incentive is to extend rather than conclude engagements; labor crew sizes and work rules that have no analogue in peer countries; and a utility relocation process in which Con Edison and other entities treat the project as an opportunity to extract maximum concessions rather than as a public infrastructure effort they are obligated to facilitate.
Eric Goldwyn, head of NYU's Transit Costs Project, has documented a structural pattern in American public infrastructure: government agencies have become "weirdly hostile to accepting the fact that there is even a problem." Janno Lieber, chief executive of the MTA, publicly dismissed researchers concerned about transit costs as "a subculture that gets a lot of their cost information from the internet." This is the institutional immune response — the defensive posture of an organization that has normalized dysfunction to the point where dysfunction no longer registers as a problem.
The Anatomy of Failure: Six Causes, One Diagnosis
I. Environmental Review as Veto Instrument
Environmental review laws — the National Environmental Policy Act at the federal level, California's Environmental Quality Act at the state level, analogous regimes across the EU — were enacted in good faith to prevent genuine environmental harm from unconsidered development. They have evolved into something else: an adversarial legal mechanism that any party can deploy to delay or kill a project they oppose for any reason, with no requirement to demonstrate genuine environmental concern and no penalty for frivolous challenge. CEQA requires responses to every public comment submitted during review. A well-organized opposition can generate thousands of comments requiring individual responses, extending review by years. Litigation follows at each decision point. The legal fees, delay costs, carrying costs on financing, and institutional knowledge loss accumulated over a decade of review and re-review dwarf the cost of most environmental mitigations that would actually have been required. The Chenab Bridge required environmental clearances. It did not provide a legal mechanism for indefinite obstruction.
II. Procurement as Risk Theater
Modern Anglo-American public procurement has evolved toward contracts that attempt to transfer maximum risk to private contractors. The theory is sound — taxpayers shouldn't bear construction risk. The practice is catastrophic. Contractors respond to maximum risk-transfer by pricing that risk into bids, then fighting expensive change-order battles when reality diverges from the specification, which it always does on complex novel projects. HS2 Ltd's contracts with its four major construction contractors were structured on precisely this model; HS2 Ltd "did not heed the warnings" of the Public Accounts Committee, as the committee noted acidly in 2024, that bearing more of the risk would require more careful contract management — which HS2 Ltd did not provide. The result was adversarial relationships, contested change orders, cascading delays, and a renegotiation process that the committee was "unconvinced" would produce meaningful savings. The Hoover Dam's Six Companies consortium had a clear scope and an owner that knew what it wanted. The USBRL used a design-and-build model with an engaged, technically capable owner-representative. Neither would survive modern Western procurement rules.
III. Hollowed-Out Public Technical Capacity
Research by Zachary Liscow of Yale, Will Nober of Columbia, and Cailin Slattery of UC Berkeley, published at the Brookings Institution's 2024 Municipal Finance Conference, found that limited state Department of Transportation capacity is one of the two largest drivers of American infrastructure costs. States that invest in DOT engineering staff have significantly lower project costs — a one standard deviation increase in DOT employment per capita correlates with 16 percent lower construction costs. Most state DOTs have been systematically understaffed over decades, replacing engineers with management consultants who have incentives aligned with project extension rather than project completion. When a consultant is billing by the hour and the scope is unclear, the consultant studies it further. The state engineer would have resolved it in a meeting.
The Eno Center's Paul Lewis put it directly: "A lot of the costs of managing tunnel projects comes down to the sheer capacity and expertise of public sector staff; we, as a country, tend not to invest enough in that." The same dynamic is visible in Britain: the Stewart Review found that HS2 Ltd pushed forward on schedule "before there was sufficient design maturity" — the organizational equivalent of starting construction without knowing what you're building, a failure mode that would not have been possible with an adequately staffed and technically competent owner.
IV. Gold-Plating and Scope Creep
The Stewart Review's finding that HS2's "top-down vision of building a railway that would be the best and fastest" destroyed cost control is a precise description of a pathology common across Western megaprojects. California HSR has been redesigned multiple times to accommodate political demands — additional stations, revised alignments, aesthetic standards — each individually defensible and collectively devastating. New York's Second Avenue Subway stations are, by the analysis of the Transit Costs Project, functionally three times larger than they need to be. The mezzanines above the platforms exist primarily because someone decided the stations should be architecturally significant. They cost hundreds of millions of dollars that Madrid and Seoul did not spend, because Madrid and Seoul built stations sized for function rather than statement.
The Chenab Bridge was designed to a specification and built to that specification. When the arch span was revised from the original design to 467 meters during the 2009 redesign, it was because the engineering required it — not because a stakeholder demanded a different aesthetic.
V. The Legal Structure of Obstruction
Leah Brooks of George Washington University and Zachary Liscow of Yale, in a landmark 2019 paper tracing the tripling of real per-mile interstate highway costs between the 1960s and 1980s, identified a precise inflection point: the wave of citizen-participation and environmental legislation enacted around 1970 — NEPA, CEQA, the Supreme Court's 1971 Citizens to Preserve Overton Park v. Volpe decision limiting judicial deference to agency decisions. Before 1970, local income had no statistically significant relationship to per-mile highway costs. After 1970, the correlation quintupled. Wealthier communities acquired effective legal tools to make infrastructure more expensive and slower — tools that have been refined and extended over fifty years into the comprehensive obstruction apparatus that now characterizes American and British project development. Brooks and Liscow do not argue that citizen participation brought no authentic benefits. They argue that the legal structure evolved beyond its original purpose into something that functions primarily as veto power for those with resources to exercise it.
VI. The Contractor Concentration Problem
Liscow, Nober, and Slattery found a second major driver of American infrastructure costs: declining contractor competition. Most states have fewer construction firms than a decade ago. Most state DOTs do little bidder outreach. An additional bidder on a project is associated with 8.3 percent lower costs; a 12 percent increase in outreach is associated with 17.6 percent lower costs. The U.S. construction industry has consolidated to the point where most large projects face limited competitive pressure on bids, further inflating costs that are already elevated by every other pathology in the system.
| Project | Country | Cost / km | Status |
|---|---|---|---|
| China HSR network average | China | $17–21M | 48,000 km operational |
| USBRL incl. Chenab Bridge | India | ~$19M | 272 km, inaugurated June 2025 |
| Spain HSR average | Spain | $20–30M | 4,000+ km operational |
| Europe HSR average | EU | $25–39M | Varies by line |
| California HSR (Central Valley IOS) | USA | ~$100–160M | 119 mi under construction; 0 revenue service |
| HS2 Phase 1 (London–Birmingham) | UK | ~£595M (~$760M) | Under construction; 2036–39 projected opening |
| NYC 2nd Ave Subway Phase 2 (projected) | USA | $3.2B | Not yet contracted |
What Countries That Are Still Building Do Differently
China's cost advantage is partly authoritarian — land acquisition that requires a decade of litigation in California can be compelled in months when the state can override property rights summarily. That explanation, however, only goes so far. Spain, South Korea, Turkey, and France are genuine democracies with rule of law, property rights, and active civil societies. They build high-speed rail for between one-fifth and one-tenth the per-kilometer cost of the United States. They do not do so by paying poverty wages; European construction wages are broadly comparable to American ones. They do so because their institutional structures prioritize project completion over process perpetuation.
India is the most instructive case because it is a genuine democracy — noisy, litigious, politically complex — that managed to build the world's highest railway bridge through a disputed Himalayan territory, in Seismic Zone V, while facing active insurgency and documented terrorist threat, at a cost that makes California HSR look like an accounting error. The instrument that made this possible was the declaration of USBRL as a National Project. This was not merely a funding mechanism. It was a signal to every institution in the Indian government that this project was necessary and that obstacles would be treated as engineering problems to solve, not as legitimate permanent vetoes to be litigated indefinitely. The Konkan Railway Corporation as owner-representative maintained technical competence and institutional authority throughout construction. The design-and-build contract structure aligned contractor incentives with project completion. The BIM requirement, mandated from inception, eliminated the information asymmetry between designer, contractor, and owner that fuels Western change-order wars.
The World Bank, in its analysis of China's HSR cost advantage, identified standardization as the primary mechanism: standardized track designs, standardized viaduct beam casting, standardized rolling stock procurement, standardized signal systems — all produced at scale by state-owned enterprises optimized for volume rather than margin. The per-unit cost of a standard HSR viaduct beam in China is a small fraction of the equivalent in the United States not because Chinese workers are cheaper but because China has built enough of them that the design, tooling, and construction process have been optimized to an industrial standard. The United States, by contrast, treats every infrastructure project as a bespoke creation requiring fresh design, fresh environmental review, fresh procurement, and fresh legal challenge.
The Historical Discontinuity
What makes the current situation particularly troubling is that the United States demonstrably possessed the institutional capacity to build large things quickly within living memory. The transcontinental railroad: six years. The interstate highway system: authorized 1956, substantially complete by the mid-1970s, 46,000 miles at costs that, even adjusting for inflation, are a fraction of current per-mile figures. Hoover Dam: five years, under budget. The Empire State Building: thirteen months. The Pentagon: sixteen months.
Brooks and Liscow's research establishes that the cost tripling in interstate construction between the 1960s and 1980s is not explained by higher wages, more complex projects, or more sophisticated specifications. It is explained by the legal and institutional changes of roughly 1970–1975. The institutional capacity to build at the pace and cost of the postwar era has not been lost to some irreversible technological or economic change. It was legislated away, precedent by precedent, regulation by regulation, over fifty years. In principle, it could be legislated back.
In practice, the political economy of doing so is extraordinarily difficult. Every element of the obstruction apparatus has constituencies that benefit from it: the lawyers who file CEQA challenges, the consultants who conduct environmental reviews, the contractors who profit from change-order disputes, the communities who use process to enforce their preferences on infrastructure routing. None of them will vote for their own disempowerment. The United Kingdom's Transport Secretary called HS2 "an appalling mess" in June 2025 and committed to reform. California's High-Speed Rail Authority released a supplemental report in August 2025 proposing revised engineering standards to lower costs. These are genuinely encouraging signals. They are also, at this point, several decades overdue.
"We seem to miss the mark on the governance piece. The institutions we have that build transit are not always adequately prepped and able to do that — from a board standpoint, from a structure standpoint."
Paul Lewis, Policy Director, Eno Center for TransportationThe Strategic Dimension
The infrastructure gap between the West and its strategic competitors is not merely an economic inconvenience. China's 48,000 kilometers of high-speed rail connects economic centers and regional cities at a pace and scale that no Western nation has matched in the postwar era. India's USBRL — whatever its geopolitical dimensions — delivers permanent all-weather military logistics access to the Kashmir Valley and the Line of Control, capability that no road can reliably provide. Russia's Kerch Strait Bridge, damaged and repaired multiple times since 2022, continues to supply forces in Crimea because the strategic imperative to keep it operational outweighs any construction difficulty in the Russian decision-making framework.
The United States has not built a single mile of true high-speed rail. The United Kingdom is struggling to complete 135 miles of it, at a cost that defies rational description, by the late 2030s. Meanwhile, the Pentagon spends billions annually on logistics infrastructure in contested regions that a reliable rail network could replace. The American Society of Civil Engineers grades U.S. infrastructure overall at C-minus. Bridges rated structurally deficient number in the tens of thousands. The gap between what the United States needs to build and what it is capable of building, at current institutional capacity and current cost trajectories, is not closing. It is widening.
The transcontinental railroad was built because the United States decided it was necessary and removed the obstacles in its path. The Hoover Dam was built because the United States decided it was necessary and built it. The Chenab Bridge was built because India decided it was necessary and built it. The question is not whether the West knows how to build. The question is whether it still knows how to decide.
Comments
Post a Comment