Counterargument

On what’s already being tried

Doesn’t outcomes accountability already exist? Federal performance measures under MAP-21 and the FAST Act require states to report on safety, pavement condition, and other targets. It exists, unevenly, and the unevenness is itself revealing. For safety specifically, there’s a real mechanism: states set their own annual targets against five federal measures, and a state that misses most of them must submit an improvement plan and is capped at the prior year’s safety-funding level for one year. That’s a genuine, if narrow, consequence. For environmental sustainability, there’s essentially nothing comparable — states and regional planning organizations set their own emissions-reduction targets, and federal regulation is explicit that these targets are neither approved nor rejected by anyone. No review, no consequence, no loop.

That unevenness is the more precise version of this critique: it’s not that outcomes accountability is uniformly absent, it’s that its strength varies dramatically by outcome, with no evident logic connecting which outcomes get real institutional teeth and which get none. Stephanie Pollack’s work at MIT’s Mobility Initiative has documented this measurement landscape in detail — her research on evaluating state DOT effectiveness is the clearest existing account of what states currently track. This argument doesn’t compete with that work; it picks up where it stops, at the institutional question of why the follow-through mechanism exists for some outcomes and not others, and what it would take to make it consistent.

Isn’t this just project-level conditionality by another name — tying funding to conditions, which states already do? Project-level conditionality (build X, in exchange for funding Y) is common and useful, but it’s a different mechanism than what’s being proposed here. It governs a single project’s construction, not a jurisdiction’s ongoing performance against a goal it has already committed to. The model closer to what’s needed is RHNA’s enforcement chain on the housing side: state-level targets, mandatory certified plans, performance-triggered consequences, and eventually a private right of action when plans go unmet. That’s system-level and cyclical, not project-level and one-time.

States already track ridership, VMT, and mode share. Isn’t that outcomes accountability? Tracking a proxy isn’t the same as tracking an outcome. Ridership, vehicle miles traveled, and mode share are measurements that may or may not indicate progress toward what actually matters — accessibility, safety, sustainability, affordability. Rising ridership doesn’t prove people can reach more of the destinations they need; falling VMT doesn’t prove anyone’s better off. Most states that report these numbers have never closed the loop back to the higher-order outcome the proxy was supposed to indicate. That gap — reporting a proxy without ever testing what it means for the real goal — is a large part of what this argument is trying to fix.

On whether this is even measurable

Housing has a clean metric: units built. Transportation doesn’t have an equivalent. Isn’t that why outcomes accountability is harder here? It’s genuinely harder, but not for the reason usually assumed. The difficulty isn’t that transportation lacks measurable outcomes — Brookings’ 50-state survey shows states already collecting meaningful data in pockets, on safety, on emissions, on access. The difficulty is that transportation has multiple, sometimes competing higher-order outcomes rather than one legible number, which makes institutional follow-through harder to design, not impossible to design. That’s a reason to build a more careful mechanism — one that can hold a jurisdiction accountable to whichever outcomes it commits to — not a reason to give up on mechanism-building altogether.

On whether this backfires

Won’t this just become another compliance burden — more reporting requirements, without any more actual delivery? That’s a real risk, and it’s the same risk Jennifer Pahlka has documented across digital government reform as the “cascade of rigidity” — well-intentioned accountability requirements calcifying into checklists that measure compliance with a process rather than progress toward a goal. The distinction that matters here is between a reporting requirement and a learning cycle. A static report card that dead-ends after publication is compliance. A cycle that requires diagnosis of why a gap exists, a tested corrective response, and a documented lesson fed back into the next round is something categorically different — closer to a feedback loop than a filing requirement.

Aren’t transportation agencies already trying to do this — they’re just underfunded and understaffed? Often, yes — and that’s a real constraint, not a strawman. But underfunding and the absence of an institutional mechanism are two different problems, and more money alone doesn’t solve the second one. An agency can be fully staffed and well-funded and still have no structure that connects a stated goal to an investment decision to a measured result. The argument here isn’t that agencies or the people in them have failed — it’s that the system rarely asks them to close this particular loop, and building the tools to ask is separate work from building the budget to answer.

On the abundance movement’s own blind spots

Isn’t this just the abundance movement’s blind spot showing? David Zipper has argued the movement’s cost/delivery framing ignores whether the things it builds actually work. Zipper’s critique is largely right, and this argument is meant as a direct answer to it rather than a defense against it. If the abundance movement’s transportation wing stops at “build faster and cheaper,” it hasn’t actually delivered on abundance’s own promise — more of what people need, not just more construction. The response isn’t to abandon the cost/delivery work Zipper is critiquing; it’s to add the piece his critique correctly identifies as missing: an accountability structure connecting delivery to the outcomes it was supposed to produce.

Improved land use — more density, more transit-oriented development — will deliver transportation outcomes on its own. Density genuinely matters — Zipper is right that living closer to destinations shortens trips and lowers the cost of getting around, and no version of this argument is against building more housing near transit. But Alex Armlovich’s related point about transit-oriented development exposes the gap in the density-first argument: TOD only delivers on its promise if the transit it’s built near is actually part of a functioning, interconnected, reliable network — and most often, it isn’t. From a land-use perspective, density is unambiguously good policy. From a transportation-outcomes perspective, density is an input — arguably the most important input, but still an input, not the outcome itself. Building density near a transit line doesn’t guarantee anyone can live car-free or car-light unless the transit, walking, and biking systems around that density are actually built and maintained to function. That’s the same accountability gap this whole argument is about, just one layer closer to land use: a jurisdiction can build all the density it wants and still never measure, or be held to, whether the surrounding system actually lets people leave the car at home.

On other proposed fixes

High-speed rail is what’s actually needed. Shill and Levine address this directly in Transportation for the Abundant Society: abundance’s transportation discourse has focused disproportionately on high-speed rail and megaprojects, but HSR is largely irrelevant to the everyday task of moving within a metro area — which is where the vast majority of trips, and the vast majority of the accessibility problem, actually happen. Even its best-case benefits are concentrated in a handful of corridors, and there aren’t many additional places in the U.S. where new HSR is realistic. The fixation on HSR as abundance’s transportation answer has crowded out attention to the intracity transportation and land-use problems that determine whether people can actually reach the destinations they need — which is the harder, more consequential, and more widely applicable fight.

What we actually need is better cost/delivery — fix the unions, fix the procedural fetish. This is the argument the Niskanen piece is written to directly answer — see Infrastructure Delivery Is Not Outcomes Delivery for the full case. In short: cost and delivery reform is necessary but not sufficient. Even solved perfectly — Swiss or Japanese-level construction costs — the evidence doesn’t support the assumption that outcomes follow automatically. Highway widenings get more congested; rail systems built at any cost still lose riders if the accountability infrastructure connecting investment to outcomes doesn’t exist. This isn’t an argument against cost/delivery reform. It’s an argument that cost/delivery reform alone has never been shown to produce the outcomes it’s assumed to produce.

Autonomous vehicles will deliver these outcomes. Possibly, for one outcome — AVs may plausibly reduce crash rates by removing human error, though the evidence is still developing and often overstated. But safety is one higher-order outcome among several, not a substitute for the others. AVs don’t scale well in dense urban environments and are unlikely to improve accessibility in the places accessibility is hardest to achieve today. And AVs are, almost by construction, an expensive way to move one person — nothing about the technology addresses affordability; if anything, it likely worsens it relative to transit. Treating AVs as a general answer to “outcomes” mistakes a possible safety improvement for the whole higher-order-outcomes framework.

This is too big. Let’s do something more manageable, like e-bikes and micromobility. Micromobility is genuinely good policy, and nothing here argues against it. But Brian Hanlon’s discipline with California YIMBY is instructive here: a coalition that lets itself get pulled into every adjacent, easier-to-win cause dilutes the one lever that actually moves the outcome. Micromobility investment is valuable on its own terms; it is not a substitute for building the institutional mechanism that connects any transportation goal — including micromobility’s own stated goals — to measured results. Somebody has to do the harder, structural work. That doesn’t mean nobody should do the easier, adjacent work too.

On political feasibility

The current presidential administration is hostile to anything non-car. True, and worth taking seriously rather than waving away — but it’s a smaller obstacle than it appears. The vast majority of transportation decision-making — project selection, service design, capital allocation — happens at the state and local level, largely insulated from federal executive posture. And the deeper point cuts the other way: car-centric transportation policy is also failing to deliver on its own stated higher-order outcomes. Federal hostility to transit doesn’t produce safer roads, lower costs, or better access by car either. The accountability gap this argument identifies isn’t partisan — it shows up in red states and blue states alike, in car-first policy and transit-first policy alike. Jennifer Pahlka has made a version of this case explicitly: rebuilding government’s capacity to deliver can’t be a partisan project, because voters can’t reward results they never actually see delivered — a problem for everyone in office, regardless of party, not a lever either side can use against the other.

We just need to win elections and get the right people in charge. Pahlka’s central frustration, developed across her writing and interviews since Recoding America, is precisely that American political culture over-focuses on policy and politics and under-focuses on implementation — winning the argument, winning the office, passing the bill — while treating what happens after as a technical afterthought rather than the actual site of success or failure. Her point isn’t that politics doesn’t matter. It’s that a movement that stops at “get the right people elected” hasn’t finished the job; it’s just gotten to the starting line. The right officials, without institutional mechanisms that outlast any single administration, produce the same undelivered promises the next election cycle.

On whether people actually want this

Americans don’t actually want these outcomes. Already addressed directly in Seattle’s Transportation Goals Are Strong. It’s Missing the Follow-Through. [link pending] — the short version: this is precisely the kind of cynicism the abundance movement exists to push back against. Seattle’s Transportation Plan passed unanimously after two years of public engagement, with advocates pushing for more ambition, not less. That’s not the behavior of a public that doesn’t want the outcome. Assuming insincerity when a democratically adopted goal goes unmet isn’t a neutral read of the evidence — it’s an excuse not to ask what’s actually standing in the way.

Fine, Americans want these outcomes, but they don’t want to change their behavior. This assumes the choice being tested is real, and it usually isn’t. Travelers, like any customers, choose among products based on each product’s competitive advantages relative to its alternatives. Right now, government transportation policy — road design, parking mandates, signal timing, capital allocation — overwhelmingly prioritizes the car product, often in direct contradiction to the jurisdiction’s own stated goals. That’s not a revealed preference experiment; it’s a rigged one. Before concluding that Americans have chosen cars and don’t want alternatives, the honest test is to see what happens when other products are actually allowed to compete on their own merits — safety, cost, reliability, time — rather than starting the race with one competitor’s thumb already on the scale.