Mega projects usually begin with a beautiful lie. Not always a dishonest lie. Sometimes it is more subtle than that. A clean line on a map. A new rail corridor. A bridge across water. A tunnel under a city. A port expansion. A pipeline. A new airport. A grand national scheme presented with polished renderings and neat numbers.
The public sees ambition. Politicians see legacy. Investors see growth. Consultants see contracts. Local authorities see regeneration. Everyone gets something from believing the project will work.
I understand the attraction. There is something deeply human about wanting to build big things. A country that cannot build begins to feel tired. A city that cannot expand begins to feel trapped. Infrastructure is confidence made physical. It says: we are still capable of doing difficult things.
But the problem is not ambition. The problem is when ambition escapes geography.
That is where many mega projects begin to fail. Not in the press conference. Not in the funding announcement. Not even in the engineering drawings. They fail when the proposed corridor is treated as a political idea rather than a spatial reality.
A corridor looks simple from above. A line from A to B. That is how major projects are sold because the public understands lines. London to Birmingham. City centre to airport. Mine to port. Oilfield to refinery. Inland city to coastal terminal.
But no corridor is just a line. It is a sequence of constraints. Land ownership. Elevation. Slope. soil condition. flood risk. protected habitats. existing utilities. settlements. roads. rail crossings. rivers. archaeological sites. political boundaries. construction access. maintenance access. emergency access. noise exposure. visual impact. compensation claims.
The line is the easy part. The corridor is the problem.
I think this is one of the most common planning mistakes in infrastructure. The project team falls in love with a strategic line before understanding the messy geography beneath it. Once that happens, every later stage becomes a defensive exercise. Instead of asking, “What is the best route?” the project starts asking, “How do we justify the route we have already chosen?”
That is a dangerous shift. It turns spatial analysis from a decision tool into a decoration.
Mega projects fail with almost comic regularity. Bent Flyvbjerg’s work on major projects has become famous partly because the pattern is so brutal: around nine out of ten mega projects go over budget, a figure also cited in industry analysis by McKinsey. 
That should stop people in their tracks. Nine out of ten. Not one or two embarrassing exceptions. Not a few unlucky cases. A structural pattern.
The Boston Big Dig is one of the classic examples. It was originally expected to cost around $2.8 billion and be completed in the late 1990s. It was eventually completed years later at roughly $14.6 billion, after delays, leaks, design issues, legal disputes, and public anger. 
California High Speed Rail tells a similar story in a different setting. The project was originally sold to voters with a far lower cost estimate than today’s projected range. Reuters reported in 2025 that the project, initially estimated at $33 billion with completion expected in 2020, had risen to an estimated cost of between $89 billion and $128 billion. 
Britain’s HS2 became another warning sign. The northern leg to Manchester was cancelled in 2023, and a UK Public Accounts Committee report later noted continuing disagreement between government and HS2 Ltd over the cost of completing Phase 1, with inflation-adjusted programme costs possibly approaching £80 billion. 
These projects are different. Different countries. Different political systems. Different terrain. Different procurement structures. Yet the same pattern keeps appearing. Costs rise. Timelines slip. Scope changes. Public trust collapses. And then everyone behaves as if the outcome was unforeseeable.
I do not think it was unforeseeable. I think the warning signs were spatial long before they were financial.
A poor corridor choice does not produce one clean failure. It produces dozens of small failures that accumulate.
A route cuts through difficult ground, so tunnelling increases. It crosses dense urban areas, so land acquisition becomes expensive. It passes near protected environments, so permitting slows. It avoids one politically sensitive area but creates engineering complexity somewhere else. It underestimates utility relocation. It assumes construction access where access is actually poor. It ignores flood zones, unstable slopes, or contaminated land until later surveys expose the reality.
By that stage, the project is trapped. Too much political capital has been spent. Too many contracts have been signed. Too many public promises have been made. The route is no longer a technical decision. It is an institutional commitment.
That is when costs become almost impossible to control.
The public usually sees the problem as waste. Sometimes it is. But often the deeper issue is that the original spatial assumptions were wrong. The project was priced as if the corridor were simpler than it was. The final cost is not just inflation or mismanagement. It is reality arriving late.
And reality is expensive when it arrives after approval.
Mega projects often fail because the politically attractive corridor is not the functional corridor.
A politician wants a route that connects certain cities. A regional authority wants a station in its district. A developer wants land value uplift. A ministry wants symbolic national integration. Local stakeholders want access but not disruption. Environmental groups want avoidance. Communities want benefits without noise, compulsory purchase, or construction chaos.
The final corridor becomes a compromise. That is normal. Infrastructure exists in society, not in a laboratory. But when compromise overwhelms spatial logic, the project starts carrying hidden costs.
A rail line may bend to serve political priorities, reducing speed and increasing engineering complexity. A road may avoid one settlement only to cross worse terrain. A pipeline may take a longer route to avoid permitting conflict but increase long term maintenance exposure. A port access corridor may be designed around current traffic while ignoring future logistics growth.
This is where spatial analysis becomes uncomfortable but necessary. It shows the trade offs clearly. It reveals who benefits, who pays, and where risk is being displaced. That is why some people do not like it. Good GIS does not flatter political convenience. It exposes it.
There is another reason mega projects fail. Early numbers are often too optimistic because they need to be.
A realistic number can kill a project before it begins. So the first estimate is softened. Contingencies are narrow. Risk is described as manageable. Land acquisition is assumed to be smooth. Public opposition is underestimated. Geology behaves itself. Utilities are where the records say they are. Environmental approvals move through the system. Construction markets remain stable. Labour is available. Materials do not spike. Political support does not fracture.
This is not planning. It is hope wearing a suit.
I think this is especially dangerous in corridor projects because risk compounds over distance. A single site can be investigated deeply. A corridor stretches across many conditions. Every kilometre introduces new uncertainty. If the early route assessment is too shallow, the project carries that ignorance forward like a hidden debt.
Eventually the debt has to be paid.
Spatial analysis cannot remove uncertainty, but it can price it more honestly. It can identify high-risk sections before they become contractual disasters. It can compare route options using terrain, land use, population exposure, hydrology, environmental sensitivity, construction access, and asset interface data. It can show where the project is likely to bleed money.
That is exactly why it should happen early. Not after the route has become politically sacred.
Delay is often discussed as a management issue. Sometimes it is. But delay also has geography.
A project may be delayed because one section requires complex tunnelling through unstable ground. Another section may be delayed by land disputes in a dense urban district. Another may be held up by flood defence redesign. Another by habitat mitigation. Another by archaeological discoveries. Another by road closures needed for utility relocation.
These are spatial delays. They occur in places for reasons tied to place.
When project managers treat delay as a generic timeline problem, they miss the pattern. They see slippage but not geography. A spatial view allows delays to be mapped as risk clusters. It shows whether the problem is land, terrain, regulation, access, utilities, communities, or environmental exposure.
That matters because each type of delay requires a different response. You cannot solve geology with stakeholder engagement. You cannot solve land acquisition with better drainage. You cannot solve protected habitat constraints with faster procurement.
The map tells you what kind of problem you actually have.
Mega projects do not only suffer cost overruns. They also suffer benefit shortfalls. Passenger numbers fail to meet forecasts. Freight demand is lower than expected. Economic regeneration is concentrated in fewer locations than promised. New roads induce congestion rather than relieve it. New stations raise land values but do not transform productivity. Industrial zones are built without enough market depth.
This is another spatial failure.
Demand is not evenly distributed because people and businesses do not behave evenly across space. A transport corridor may connect two cities but miss the real commuter flows. A logistics hub may sit near a highway but too far from labour pools. A new airport may be technically impressive but poorly integrated with urban transport. A retail-led regeneration scheme may assume catchment strength that does not exist.
The mistake is treating connectivity as automatic value. It is not. Connectivity only creates value when it matches actual movement, demand, land use, and economic behaviour.
I think this is where many benefit-cost models become too abstract. They convert geography into assumptions. Then the project fails because the assumptions were never spatially tested in enough detail.
Good spatial analysis does not guarantee success. Nothing does. But it improves the honesty of the conversation.
It can reveal route sections where slope, soil, flood risk, or geology increase construction complexity. It can identify communities facing disproportionate disruption. It can show where environmental constraints are likely to trigger delays. It can expose weak access points for construction logistics. It can identify utilities and existing assets that create relocation risk. It can compare corridor options using weighted criteria rather than political instinct.
Most importantly, it can show decision makers the price of pretending.
A proper corridor analysis might say: this route is politically attractive but technically fragile. This alternative is less glamorous but more buildable. This station location looks good on a campaign leaflet but performs poorly in catchment analysis. This route saves distance but increases tunnelling exposure. This alignment avoids one public objection but creates a larger environmental problem.
That is the kind of intelligence mega projects need before the big announcement, not after.
Mega projects fail because institutions often make promises before they have fully understood place.
That is the uncomfortable truth. The failure is not only engineering. It is not only procurement. It is not only inflation. It is not only politics. It is the gap between the imagined corridor and the real corridor.
The imagined corridor is smooth. The real corridor is contested, uneven, wet, crowded, unstable, protected, owned, regulated, buried with utilities, and full of people who may not want the project near them.
The imagined corridor lives in a presentation. The real corridor lives in mud, rock, law, weather, traffic, money, and politics.
This is why I think the best infrastructure planning begins with humility. Not small ambition. Humility. The willingness to let the map challenge the idea before the idea becomes too expensive to challenge.
If governments and companies want mega projects to succeed, they need to move spatial analysis to the front of the process. Not as a technical appendix. Not as a supporting map. As the foundation of decision making.
Before a corridor is announced, it should be stress-tested. Before a cost is promised, the route should be risk-weighted. Before benefits are claimed, demand should be spatially validated. Before political commitments are made, decision makers should understand where the project is likely to face land, environmental, engineering, and community friction.
This will not make every project cheap. Big infrastructure is expensive because the modern world is dense, regulated, and technically demanding. But it can make projects more honest. And honesty at the beginning is far cheaper than revelation at the end.
I still believe in mega projects. I think countries need to build. They need railways, grids, ports, flood defences, tunnels, pipelines, housing infrastructure, and water systems. A society that cannot build becomes a museum of its own former competence.
But building well starts with respecting geography.
The corridor is never just a line.
It is the project.