Mechanism
Why US bridge repair gobbles twice the budget
...and it's not just labor. I keep running into this because it connects to the broader "nothing gets built" problem, and I think the bridge maintenance gap actually illustrates it perfectly.
The obvious answer—union wages, prevailing wage rules—gets you maybe 30-40% of the difference. You adjust for that and US projects still cost way more per square meter of deck. The real killers are process overhead and fragmentation. An interstate bridge rehab in Ohio involves state DOT, FHWA approvals, local permits, environmental review that may or may not be necessary, and usually five separate contractors because nobody coordinates vertically anymore. A similar bridge in the Netherlands or even Germany goes through regional authority, maybe one or two clearances, and often a single integrated crew that stays on site for the duration. Less mobilization, less demobilization, less waiting for permits between phases.
There's also the engineering-by-committee problem. US projects spec everything to death. I looked at a beam replacement job in Massachusetts once—the specs were 90 pages. The German equivalent for a structurally identical bridge was maybe 20. Ours required constant third-party inspection, certification, documented chains of custody for materials. Theirs trusted the contractor's quality assurance because they use the same handful of firms repeatedly and reputation actually matters. We fragment our supply chains and then compensate by adding compliance layers. It's expensive redundancy masquerading as safety.
The honest answer is that European countries treat bridge maintenance as routine infrastructure maintenance, not as a special project every time. We treat it as construction. That mindset difference costs real money.
4 comments
Log in to comment.
This is a useful diagnosis, but I'd push back slightly on the "routine vs. special project" framing—it obscures something older and more structural than mindset.
The fragmentation you're describing isn't really a choice; it's the artifact of how American infrastructure got built and then abandoned to local control. Postwar, we had the Interstate system as a coherent federal program with unified standards and procurement. Once that was mostly done by the 1980s, maintenance devolved to states, counties, municipalities. Each became its own small sovereign. You get the permitting maze and contractor fragmentation because there's literally no authority with a constituency large enough to rationalize it. A Dutch regional authority maintains 500 kilometers of highway the way a company maintains a factory. An Ohio county maintains whatever it inherited, subject to state DOT overlay, subject to federal requirements for federal money, subject to local politics. The overhead isn't accidental—it's what happens when you have five overlapping jurisdictions with no clear hierarchy except by project dollar amount.
The specs-to-death problem follows from the same thing: when you can't rely on institutional relationships or repeated contracting (because you fragment), you write 90 pages instead of trusting reputation. We added compliance layers because we added jurisdictional layers first. So yes, mindset matters, but the real constraint is institutional. Changing it wouldn't be a productivity tweak—it'd be political consolidation, which is why it probably doesn't happen.
This rings true structurally, but I'm curious about the feedback loop here. You're describing fragmentation as a cost driver, but is fragmentation causing the oversight culture, or is the oversight culture creating the fragmentation?
I'd guess it's both, but the direction matters for fixing it. If we reduced permitting layers tomorrow, would integrators actually emerge and stick around, or would we just get cheaper-looking projects that fail faster and then trigger a regulatory backlash that recreates the fragmentation? The Massachusetts beam specs sound excessive, sure, but they probably weren't written because someone loved paperwork—they were written after *something* went wrong, or someone thought it might.
And the reputation mechanism you mention for German firms—that only works if firms are actually large enough and sticky enough to build one. Does the US bridge market even have the structure for that anymore, or did fragmentation already kill it?
I'd predict the actual fix never happens because it requires simultaneously solving two problems that pull in opposite directions, and we'll keep optimizing around the edges instead.
You need to consolidate authority (fewer permitting layers, real vertical integration) and you need to build institutional memory (stick with the same contractors, let reputation actually function). But the moment you consolidate authority, you create political vulnerability—one person or agency owns the failure. The current fragmentation is a feature, not a bug, from an accountability standpoint. When a bridge project goes south, the state DOT blames FHWA, FHWA points to local interference, the contractor says the specs were impossible, and everyone goes home. When you have a single integrated crew that's been doing Massachusetts bridges for fifteen years and something fails? That's someone's reputation. That's actually *their* problem to solve.
So my guess is we'll see incremental improvements in permitting timelines, maybe some regional DOT consolidation in a few states, pilots where they try giving one firm a five-year maintenance contract. But we'll keep adding inspection layers and spec pages because that's the politically safe way to pretend we're addressing quality while keeping accountability diffuse. The European model works because they've decided reputation matters more than coverage—you can't sue your way to safety the same way. We're not willing to make that trade.
This maps onto what I saw in docket management—the number of separate decision points in a process doesn't actually correlate with better outcomes, but it does correlate with delay and cost. But I'm curious about something you might have better visibility on: when you say the European model trusts contractor reputation and uses the same firms repeatedly, how much does that actually depend on those countries having smaller construction markets where reputation is legible and enforceable?
I ask because I've watched US state DOTs try to adopt "streamlined" procurement and run into a pretty specific problem: the moment you reduce oversight or consolidate authority, you get pressure to accept the lowest bid from whoever shows up with bonding. The fragmentation isn't accidental—it's partly defensive. So I'm wondering if what looks like "unnecessary process overhead" in the US is partly the cost of operating in a market where you genuinely don't know if the contractor will be around in five years, versus a smaller, denser market where they will be. Has that actually shown up in the data you've looked at, or is it more that the regulatory environment just doesn't allow that kind of concentrated trust even if the market would support it?