No Bigger by Habit, No Smaller by Ideology

A government sized to its mission, capable enough to build, and judged by whether it delivers.

Pillar 2 · A Government That Works

Summary

The argument about whether government should be big or small is the wrong argument. The right question is whether it works for the people it is supposed to serve. Government should be mission-sized: no bigger out of habit, no smaller out of ideology, strong where public capacity is necessary, limited where power could be abused, and accountable everywhere. And it should be capable, because a government can be sized correctly on paper and still fail if it cannot decide, cannot execute, and never reviews what it is already doing.

This brief proposes five things aimed at both failures at once: a National Government Maintenance Act that puts every major program on a regular review schedule, a fix for the vetocracy that gives people a voice without giving every objection a veto, public capacity built through staffing and modern systems rather than bloat, a Tax Code Public Value Test applied to the roughly $2.2 trillion a year (FY2025) that moves through the tax code unexamined, and a government responsive enough to be judged on delivery, with the workers who deliver protected through every change.

The Principle

But being the right size is not enough. A government can be sized correctly on paper and still fail, because it cannot execute, cannot move, and never reviews what it is already doing. So this pillar is about capability. Capable means it has the people, the tools, and the authority to do hard things. Nimble means someone is actually in charge, can use the power the law already grants, and can decide without waiting years for permission. We have spent a long time arguing about size while letting capability decay. This is about rebuilding it.

The System Failure

America used to build. It built the world’s greatest rail network, a national electrical grid, the interstate highways, Social Security, the Apollo program. Today, on the things that touch people every day, housing, clean energy, transit, we feel stuck. The failure has two layers, and they sit on top of each other.

The first is that we have quietly made it almost impossible to decide anything. Over the past half century, in a well-intentioned reaction against the abuses of unchecked power, we diffused the authority to act, pushing it down and out, until we built what Marc Dunkelman calls a vetocracy: a thicket of overlapping reviews, court challenges, and local veto points in which nearly anyone can stop a project and no one is truly empowered to move it forward. We have, in his phrase, so many checks that no one is actually in charge. A broadly beneficial project can be killed by a handful of narrow objections, and the result is a government that looks powerful on paper and is paralyzed in practice. This is not an abstraction. Full federal environmental impact statements take about four and a half years on average, highway reviews over seven, with some stretching past two decades, and the mere risk of years of delay and litigation shapes what gets attempted at all. The honest version matters here, because credibility depends on it: fewer than one percent of these reviews are full statements, most projects clear far faster, and the government wins most of the lawsuits. The problem is not that we review everything to death. It is that the biggest, most important projects, the transmission lines, the rail, the clean energy, are exactly the ones that get caught, and the cost of a single six-year delay across the major infrastructure sectors has been estimated in the trillions.

The second failure is that even when we decide to build, we often cannot execute, because we have hollowed out the public capacity to manage the work. Look at California high-speed rail. It was sold to voters in 2008 at $33 billion, with trains running by 2020. Today the estimate runs between $128 and $231 billion, as much as seven times the promise, roughly $15.7 billion has been spent over sixteen years, and not one mile of high-speed track has been laid. Some of that is the vetocracy, the land that could not be assembled, the utilities that could not be relocated, the review conducted before the route was even chosen, the two full weeks the project was halted because a vehicle struck an endangered snake. But much of it is raw capacity failure: the state broke ground in 2012 with the design only fifteen percent complete, before it owned the land or finalized the route, partly to hit a federal spending deadline, and it managed the contracts and the risk so poorly that the overruns were, in the state auditor’s word, predictable. Neither failure is solved by the old fight over size. A bigger version of a government that cannot decide or execute is not the answer, and neither is a smaller version that has been starved of the people who do the work.

There is a counter-example, and it comes from a Republican-led suburb of Atlanta. Sandy Springs incorporated in 2005 as Georgia’s first new city in decades, and with only months to stand up a government from nothing, it outsourced nearly everything but police and fire, one giant contract covering public works, finance, courts, parks, and IT. For years it was the world’s most famous experiment in privatized government. Then the market changed, the bids came in high, and in 2019 the same city ran the numbers and brought most of those services back in-house, projecting about $13.5 million in savings over five years. By the city’s own accounting, the actual savings came in near $26 million, almost double the estimate. The lesson is not that outsourcing is wrong or that public delivery is sacred. It is that Sandy Springs did exactly what this pillar demands of the whole federal government: it treated the structure of government as a tool rather than a creed, measured what delivered, and changed course when the evidence said to. Outsourcing served a startup city that needed a government in months. Insourcing won when the vendors stopped delivering value. No bigger by habit, no smaller by ideology, proven in one Georgia city’s own books.

The Proposal

Five things, aimed at both failures at once.

Maintain what we run. Pass a National Government Maintenance Act. Create a regular, scheduled review of the major parts of the federal government: large programs, agency functions, regulations, tax preferences, subsidies, grant systems, and the public-facing processes people actually deal with. The standard is simple. If a program works, strengthen it. If it is outdated, modernize it. If it duplicates another, consolidate it. If it mainly produces paperwork or protects a narrow interest, phase it out. Nothing today forces this honest look, so the default is always to add and never to trim. Maintenance is not an attack on government. It is how anything built to last gets cared for.
Fix the vetocracy. Give people a voice, not a veto. This is the reform the standard platform ducks, and it is the one that unlocks the build agenda, the climate agenda, and the housing agenda all at once. The goal is a system where someone is clearly in charge of a decision, holds a real mandate, weighs the trade-offs in the open, and is accountable for the result, rather than a process where authority is so scattered that no one can be blamed and nothing can move. Concretely: name a lead decision-maker for major projects, consolidate overlapping federal reviews into a single process on a firm schedule, set enforceable deadlines, and preserve public input and fair compensation while ending the ability of any single objector to impose years of delay. The line to hold is exact. People should have a voice. They should not have a permanent veto over projects their own communities and the country need.

And because “voice, not veto” can be gamed from either direction, dressed up as deregulation by industry or denounced as deregulation by opponents, this platform states the test as a rule any permitting reform must pass, all six parts, before it earns support here:

First, even-handed coverage: the same fast, certain process applies to transmission, housing, transit, clean generation, and fossil projects alike; a reform that fast-tracks one category while leaving procedural weapons aimed at the others is industrial policy wearing reform’s clothes, and it fails. Second, deadlines, not lowered standards: the reform binds the calendar, decisions within firm timelines, consolidated reviews, one lead agency, while leaving the substantive environmental, safety, and labor standards themselves intact; speed comes from process, never from weakening what a project must actually meet. Third, input early, standardized, and time-capped: communities get real notice, real hearings, and a written response to substantive objections at the front of the process, in exchange for finality at the back; the trade is genuine participation for an end to participation-as-attrition. Fourth, judicial review bounded, not abolished: challenges get filed within a short statute of limitations, resolved on an expedited track, and remedies default to fix-the-defect rather than start-over, so courts still catch real illegality without becoming the delay mechanism of first resort. Fifth, the harms that concentrate get weighed where they land: cumulative local burdens, on the communities that have historically absorbed everyone else’s infrastructure, are assessed and mitigated or compensated as part of the decision, and tribal consultation and treaty obligations are a floor no streamlining touches, because a nation-to-nation obligation is not a procedural delay. Sixth, who decides is written down: the distinction between legitimate harm and obstruction is made by the accountable lead decision-maker, in a published decision responding to the record, appealable on the bounded track above, not by whoever can outlast whom.

Applied honestly, this test cuts both ways, and that is the point. It is the standard by which this platform supports permitting reform for clean energy and housing, and the standard by which it opposes the fossil-industry version that fast-tracks pipelines while leaving everything else stuck. One test, every project, no exceptions for anyone’s favorites, including ours.
Build public capacity, not bloat. The reason big projects run over is very often that the public side lacked the skilled people to manage them, the reviewers, engineers, scientists, planners, and project managers whose absence is the real bottleneck, and the modern digital systems that staff and citizens have to use. Give agencies enough of those people, fix the hiring and procurement systems that make it hard to bring them in, and modernize the technology. This is not a larger government for its own sake. It is competence, which is usually a question of staffing and skill rather than raw size. And it is the rare spending that pays for itself, because the alternative is the California pattern, projects that run billions over because no one on the public side could manage them well.
Make every tax break earn its keep. Apply a Tax Code Public Value Test. The country spends roughly $2.2 trillion a year (FY2025) through the tax code, in deductions, credits, preferential rates, exemptions, and carveouts, a sum on the order of all federal discretionary spending combined, and almost none of it is ever reviewed. Lawmakers created or modified dozens of these preferences in a single recent year alone. Every major one should have to answer a plain question: what public value does this create? Does it support work, family stability, housing supply, clean energy, domestic production, research, retirement security, small business, or mobility, or does it mainly reward those who can afford the lawyers to find it? Preferences that serve a real public purpose stay and are strengthened. The rest are reformed or ended. This is one of the strongest tools available for fiscal honesty, and it belongs both here and in the fair-markets agenda.
Be responsive, and protect the workers who deliver. Much of responsive government needs no new law. It needs leadership that uses the authority it already has, sets clear delivery goals, and is judged against them, and it means letting routine paperwork be automated so public servants can spend their time on judgment and on people. AI should handle the paperwork. People should make the decisions. And when programs are consolidated, digitized, or phased out, workers should be retrained and redeployed into the mission areas that are short-staffed, with their benefits and standing protected, rather than fired at random. Frontline workers should help design the changes, because they usually understand the system’s real failures better than anyone above them.

Implementation Pathway

The Maintenance Act, the Tax Code Public Value Test, and the worker-protection framework are federal statutes. The vetocracy reforms run partly through statute, consolidating reviews, setting deadlines, naming lead agencies, and partly through the same permitting and transmission reforms in the build and climate pillars, since this is the shared machinery underneath all of them. The capacity investments and digital modernization run through appropriations, where the staffing that determines whether government can execute is actually decided. And much of responsive government requires no new law at all, only an administration that uses existing authority well and lets itself be measured on delivery. That part can begin immediately.

Funding and Public-Value Logic

This pillar tends to save money rather than spend it. Regular maintenance catches the waste, duplication, and obsolete programs that would otherwise run forever. The Tax Code Public Value Test recovers revenue lost to preferences that serve no public purpose, out of a pool of roughly $2.2 trillion a year (FY2025) that goes almost entirely unexamined. Fixing the vetocracy saves the enormous sums lost to delay, where a single multi-year holdup on major infrastructure can cost more than the reform ever would. And capacity investment is the rare spending that pays for itself, because understaffing is precisely why large projects run over budget and behind schedule. A capable, decisive government is cheaper over time than a cheap one that cannot deliver.

Risks and Guardrails

The first risk is the one history warns about most loudly. A generation ago, unchecked builders like Robert Moses bulldozed neighborhoods and displaced the powerless, and the reaction against that abuse is what built the vetocracy in the first place. So “voice, not veto” cannot become a license to return to the bulldozer. The guardrail is precise: input, transparency, environmental protection, and fair compensation all remain, and what is removed is only the ability of a narrow interest to impose indefinite delay. The goal is a middle ground between one person deciding everything and no one able to decide anything.

The second risk is that an efficiency drive becomes a cover for gutting services people depend on, which is why the standard here is strengthen what works, not cut for its own sake, why the review must be honest rather than ideological, and why worker protection is written in. The third is that a maintenance process gets captured, so it must be insulated to serve the public rather than whichever interest most wants a given program kept or killed. And responsiveness must stay inside the law, because using authority well is not the same as stretching it past its limits.

Metrics for Success

Judge the agenda on delivery, capacity, and honesty.

Time to deliver core services and decisions, benefits, permits, approvals, and the backlog in each, tracked toward months rather than years.
Programs reviewed on schedule, and the share modernized, consolidated, or ended as a result.
Revenue and waste recovered through the Tax Code Public Value Test.
Project delivery against cost and timeline, and the staffing and skills in the roles that determine whether projects move.
Permitting and review timelines for major projects, and the share completed within their deadlines.
Public trust that government can act and deliver, measured over time.

Opposition and Responses

Some will say this is a quiet plan to cut government. It is the opposite. The standard is to strengthen what works and modernize what is outdated, worker protection is written in, and a program is only ended when it has stopped serving a public purpose, in the open. Maintenance is how you keep public institutions strong, not how you dismantle them.

Some will say government simply cannot be made to move quickly. It moved quickly for most of the twentieth century, when it built the highways and the grid, and it moves quickly today where it has capacity and where someone is clearly in charge. It stalls where authority has been scattered and where no one is accountable for delivery. Fix those two things and it can act.

Some will say government cannot solve big problems at all anymore. The ozone layer says otherwise. When scientists showed that refrigerant chemicals were tearing a hole in the sky, governments regulated them, every country on Earth ratified the Montreal Protocol, and industry innovated its way to substitutes. Nearly 99 percent of the banned chemicals are gone from production, the ozone layer is on track to heal fully within decades, and the EPA’s own accounting projects the treaty will prevent 443 million skin cancer cases and 2.3 million skin cancer deaths in the United States alone. That is what delivery looks like: a planetary problem, identified by science, fixed by regulation, solved within a lifetime. And government can deliver at speed as well as at scale. Operation Warp Speed, launched under a Republican administration and carried through under a Democratic one, put more than $10 billion behind vaccine candidates, guaranteed the purchases, ran trial and manufacturing stages in parallel instead of in sequence, and produced safe, effective vaccines in about eleven months against a process that normally takes a decade.¹ Independent modeling credits the vaccination program with averting 3.2 million American deaths² and 18.5 million hospitalizations in its first two years. The failure of some programs is an argument for maintenance, not for surrender.

Some will say “voice, not veto” tramples communities. No. Input, compensation, and environmental protection all stay. What ends is the power of a single objector to hold a needed project hostage for years. The alternative to a permanent veto is not a bulldozer, it is a fair process that actually reaches a decision, and the people most failed by paralysis are usually the ones who most needed the housing or the transit that never got built.

Some will say the Tax Code Public Value Test is a tax hike in disguise. It is a test, not a hike. Preferences that create real public value are kept and strengthened. Asking the rest to justify themselves, out of $2.2 trillion a year (FY2025) that no one currently reviews, is not raising taxes. It is ending subsidies that no longer serve the public.

What Would Change Our Position

The voice-not-veto trade is falsifiable on both sides. If streamlined review demonstrably produced the harms the old process existed to catch, environmental damage, safety failures, or projects steamrolled through vulnerable communities, the guardrails were drawn wrong and get redrawn before anything else advances. If maintenance reviews started functioning as stealth cuts, ending programs that still served their purpose, the worker and purpose protections failed and the tool gets rebuilt. And the structure-as-tool principle cuts both ways by design: where outsourcing beats public delivery on evidence, as insourcing beat outsourcing in Sandy Springs, this platform follows the evidence there too.

Public-Facing Language

Here is the short version. Stop arguing about whether government should be big or small and start asking whether it works. Build it to the size of the job, no bigger out of habit and no smaller out of ideology, and then make it capable enough to actually deliver. Put someone in charge who can be held accountable, and give people a voice without giving every objection a veto, so we can build the housing and the clean energy and the transit the country needs. Staff the government to build and decide, and automate the paperwork so public servants can serve people. Review what we already do instead of letting it run forever on autopilot, and make every tax break prove it earns its keep. A government that can act, and that people can see acting, is the whole point.

The American Mainspring Project · This printout reflects the live site at the time of printing, and updates whenever the site is updated.