During the 2016 presidential election campaign, one thing that both Democratic candidate Hillary Clinton and Republican Donald Trump agreed on was that the United States needed to renew its infrastructure. The former has proposed a $275 billion initiative to achieve just that; the latter has staked much of his claim to competence on his ability to build things. Clearly, the new President will have to do something starting in January 2017 given promises that have been made to date.

The importance of infrastructure for economic growth need not be belabored; it is part of the capital stock that enables and increases U.S. productivity. Nor has the severity of its decay gone unnoticed; the issue has been discussed extensively in the media. But the issue also bears important political implications. Building infrastructure creates large numbers of jobs for working Americans, and renewed efforts can bolster the sagging fortunes of the working class, whose anger has propelled populist politicians like Donald Trump. It is one of the few ways that the Federal government can spend money to reduce income inequality in the United States.

Infrastructure decay not only costs us economically at home, but also negatively affects our international influence and status. The United States, which used to have the finest infrastructure in the world, is now ranked 16^th according to the World Economic Forum. Corporations like Caterpillar and GE argue plausibly that the declining U.S. capacity for large-scale infrastructure hurts the nation’s ability to compete abroad.

The existing infrastructure deficit does not arise out of a vacuum: Our recent failure to build new and smarter infrastructure, as well as to properly maintain the existing stock, constitutes a microcosm of larger governance failures. There is a perfect de facto conspiracy on the part of the two main political parties to block a major push on infrastructure. The Republicans, for their part, do not want to pay for it, and have blocked increases in the gas tax that funds the Highway Trust Fund since the early 1990s. The result is that the revenue collected from that tax has not kept pace with inflation. By contrast, the Democrats feel comfortable permitting new projects to death, increasing both their cost and the length of time needed to execute a project. Between this Scylla and Charybdis, very little has gotten done.

There is an obvious and logical compromise just waiting to be sealed: Republicans should agree to new investment in infrastructure, along with the resources to pay for it. And Democrats should agree to a rationalization of the regulatory regime governing such projects. This hasn’t happened because compromise itself has become a dirty word in a polarized political environment. Whether such compromise will be possible after an angry and bitter election campaign remains to be seen.

The Constitutional Framework

Compared to other democratic political systems, and particularly to the kinds of parliamentary systems that exist in Europe, the U.S. Constitution is exceptional for the way in which it separates, divides, and indeed duplicates powers. It was of course the intention of the Framers to create a system of checks and balances that would limit the ability of a central government to tyrannize citizens. As a result, the U.S. system has many more “veto players” embedded in it than in most parliamentary systems.¹

These veto players include a separately elected President with equal democratic legitimacy to Congress; a powerful upper house of Congress whose approval is needed on virtually all legislation; an independent judiciary that can overturn legislative acts; and a complex Federal system that allocates powers to state and local governments. Most parliamentary systems have far fewer veto players, and are therefore able to pass legislation more easily. In such systems, the executive branch is not a competing power center, but rather the administrative instrument of the legislative majority.

The U.S. system has all but ground to a halt as a result of two phenomena occurring in the broader society. The first is the rise of polarization, both in the way that Americans live and interact,² and in their representative institutions, where there is virtually no ideological overlap between Republicans and Democrats today.³ The second has to do with the rise of a large number of well-resourced and professionally organized interest groups, representing everything from corporations to unions to social movements to victims of individual diseases.

When polarization and powerful organized interests combine with the U.S. check-and-balance system, we get a vetocracy, that is, rule by veto where very many players are empowered to stop the government from doing things, but where few mechanisms exist to force consensus on decisions affecting the common good.⁴ Vetocracy is also evident at the micro level, in the local decision-making regarding infrastructure projects. Such projects are ostensibly undertaken in the public interest, to provide what economists call public goods, but they inevitably hurt the interests of particular groups. Well-functioning political systems balance the general interest in public goods against the partial interests affected.

Local politics in the United States is structured in such a way as to replicate the multiple veto players of the Federal system, and adds a few more of its own. The result is that numerous affected parties have an effective veto over a large project, despite the fact that they represent only a small part of the population.

Indeed, U.S. federalism itself, at least in its current manifestation, is not organized on any principle of subsidiarity, where government functions are delegated to the lowest possible level consistent with their purpose. Powers tend to be neither separated nor delegated, but rather duplicated across branches and among different levels of government.⁵ Thus, in the area of environmental regulation, multiple Federal agencies with often overlapping jurisdictions are involved: the Environmental Protection Agency, the Bureau of Land Management, the U.S. Forest Service, the Fish and Wildlife Service, the Army Corps of Engineers, the Bureau of Ocean Energy Management, the National Oceanic and Atmospheric Administration, and the Advisory Council for Historic Preservation. Similarly, states and municipalities duplicate Federal agencies and often impose duplicative or alternative sets of regulations on projects. Thus, California has its own California Environment Protection Agency (CalEPA), empowered by its own statute, which in many instances gives it greater power than the Federal agency, as well as other state agencies with overlapping functions like the California Water Resources Control Board.

The primary mechanism through which vetocracy is exercised in infrastructure projects are the environmental reporting requirements imposed by the National Environmental Policy Act of 1969 (NEPA). NEPA requires the government to produce and make public an Environment Impact Statement (EIS) regarding any Federal action “significantly affecting the quality of the human environment.” For cases with lesser impact, the statute requires an Environmental Assessment (EA). Currently the government produces about 350 EISs and 50,000 EAs every year.⁶

No one would contest the need for the government to carefully assess environmental impacts before proceeding with large infrastructure projects. History is replete with examples of unintended consequences flowing from poorly thought-out investments in disruptive infrastructure. The writing of an EIS is, moreover, a good opportunity to invite democratic participation in the decision-making process, since NEPA requires the lead government agency to engage interested parties through public consultations, as well as to inform the public of the consequences of such projects once the decision is made.

The problem lies not in the principle, but in the inefficiency of the review process. A typical EIS requires six years to complete, at a cost of $200,00-250,000, while a typical EA takes nine to eighteen months at a cost of $50,000-200,000.⁷ Even the short-form EAs can run to thousands of pages; the review of a project to raise the Bayonne Bridge between New Jersey and Staten Island ran to 10,000 pages and took five years to complete.⁸ These reporting requirements have begotten a large industry of consultants to which much of the legwork is delegated.

The consultation process itself is both slow and costly, since it falls under a completely different set of transparency laws. These include the Freedom of Information Act (FOIA), which allows individuals to file Freedom of Information Requests; the Federal Advisory Committee Act (FACA), which requires that agency consultation with private groups be registered with the Government Accounting Office (GAO), balanced for partisan representation, and made public and open to the media; and finally, the Government in the Sunshine Act, which imposes its own, parallel set of requirements for public reporting of agency consultations. Each of these requirements slows down the review process and generates its own bureaucratic bloat, since agencies need to create special offices to deal with, for example, FOIA requests.⁹

Critics have noted many problems in the NEPA review process. The statute does not assign clear responsibility for producing the EISs and EAs; each must be reviewed by interested agencies and the reviews are often done sequentially, adding to delays. Simply choosing the “lead agency” can consume months of debate. In some cases, agencies undertake parallel and uncoordinated reviews. Failure to take account of an agency position in the initial draft sometimes leads to the document being completely re-written or blocked, or else gives rise to litigation at a later stage (see below).

Adversarial Legalism

Adversarial legalism is the term that Robert A. Kagan devised to describe the U.S. penchant for using litigation to control regulatory behavior, and at the same time to enforce regulation.¹⁰ Congress in passing the dozen or so major pieces of social legislation in the late 1960s and early 1970s, including all of the major statutes governing the environment, saw fit to broaden standing in ways that permitted a wide range of interest groups and individual citizens to sue the government, either to block unwanted regulation or to compel the government to enforce existing rules. This has led to an explosion of litigation over the years that affects agency decision-making in a variety of ways, none of them helpful in terms of regulatory efficiency. Kagan describes a case of the dredging of Oakland Harbor to accommodate larger container ships that was initiated in the late 1970s; as a result of the ensuing litigation, the project was not completed until the mid-1990s.¹¹

A similar deepening of Rotterdam Harbor in the Netherlands around the same time was accomplished in a much shorter period. European democracies do not rely on litigation to nearly the same extent as the United States; by and large, the enforcement of rules is left up to executive agencies themselves. Although the situation is changing somewhat with the rise of EU-wide courts, European national-level agencies generally are not burdened with the same level of litigation risk as U.S. ones.¹²

Judicial intervention affects the NEPA review process in several ways. The Supreme Court has blocked the use of the statute to overturn substantive agency decisions in Stryker’s Bay Neighborhood Council v. Karlen, but they can and are challenged on procedural grounds, where a large body of case law has been generated. David Hayes argues that litigation risk is not large at the Federal level; only 1 percent of Federal actions under NEPA are challenged by the courts; fewer than 175 NEPA cases have been filed annually since 2001, and the numbers appear to be declining.¹³

The consequence of judicial review in this domain is less the overturning of agency decisions than the inefficiencies that the threat of litigation imposes on the review process itself. Agencies understand that they might be second-guessed by the courts on procedural grounds, and therefore are punctilious in minimizing risk by dotting every “i” and crossing every “t” in paperwork preparation. This leads to bloated and lengthy EISs and EAs, and greatly extends the amount of time it takes to generate them.¹⁴

The State-Level Overlay

U.S. federalism today does not neatly delegate powers to state and local government as was the original intent of the arrangement; rather, it tends to duplicate powers at these levels, which adds greatly to the complexity of government decision-making. Large infrastructure projects need to pass agency review not just at a Federal level, but at a state and sometimes municipal level as well, which in a state like California can be much more burdensome than their Federal counterpart.

The California equivalent of NEPA is CEQA, the California Environmental Quality Act of 1970, which empowers the California Natural Resources Agency to oversee large projects and to disclose to the public environmental impacts through the preparation of Initial Studies (ISs) and Environmental Impact Reports (EIRs). Similar to the Federal requirements, these reports have to involve public consultation and are subject to state-level transparency regulations. It is hard to find a better example of duplicative rather than delegative federalism; virtually every Federal-level agency and procedural requirement is replicated if not amplified at the state level in California, with all of its complexity and overlapping sources of authority.

While judicial intervention and litigation may be limited at a Federal level, it vastly increases the difficulty of doing infrastructure projects in California. CEQA litigation has a number of special characteristics. First, any party can file a CEQA lawsuit; despite the fact that the underlying law was meant to protect the environment, many non-environmental groups use CEQA litigation to achieve their own ends, like seeking better terms on union contracts or disrupting business competitors.

Second, CEQA lawsuits can be filed anonymously. The transparency requirements that apply to CEQA reviews were not extended to the litigation process, so as to protect the special interests making use of it.

Third, although only 5 percent of CEQA lawsuits result in the overturning of a project, the threat of lawsuits is sufficient to significantly delay projects and raise their costs. In published CEQA appellate court cases, challengers prevailed in nearly 43 percent of suits filed. Even if they did not ultimately succeed in overturning the project, they could delay it or cause it to be altered in costly ways. Many banks or government bond issuers will not proceed with financing a project as long as there is CEQA litigation pending.

Fourth, the number of environmental topics that each study must consider has risen over time and now exceeds one hundred. Many projects aiming to mitigate environmental impacts like climate change have themselves been stymied over procedural objections in CEQA litigation.

Fifth, there is no limit to the number of times a project can be sued under CEQA. Each discretionary agency decision can be made the subject of a separate lawsuit.

Finally, while government agencies are subject to endless lawsuits, a separate California “free speech” statute prohibits “strategic lawsuits against public participation,” further protecting the anonymity of those filing lawsuits.¹⁵

The consequence of CEQA litigation is the empowerment of a wide range of political actors whose interests have little to do with environmental protection. Only 13 percent of CEQA lawsuits are filed by established environmental advocacy organizations; more than two-thirds come from business competitors, trade associations, labor unions, and, in a huge number of cases, private individuals who object to some private or public project. Surprisingly, 11 percent of lawsuits are filed by other government agencies seeking to stop another arm of the executive from doing something they oppose. Half of CEQA lawsuits target public projects rather than business development, and target “infill” projects rather than “sprawl” or “greenfield” ones. Two-thirds of all lawsuits come from individuals or families, or “other” organizations that often appear to be driven by narrow NIMBY concerns.¹⁶ We do not really know the motivations of those sponsoring these suits, however, since they are not subject to the same transparency requirements as the public agencies promoting infrastructure projects.

The state of California thus represents an extreme example of vetocracy at work, but the problem is present everywhere to a greater or lesser degree. The lack of a unified executive in the United States thanks to the evolved dysfunction of U.S. federalism means that there can be no single hierarchical point of authority that is responsible for the completion of any major infrastructure project. In theory, the legislature either at a national or state level should seek to balance the common good of public infrastructure against the individual interests affected by a given project, and authorize executive agencies to exercise appropriate discretion in the implementation stage to continue to balance those interests, based on continuous consultation with affected stakeholders. But U.S. political culture is deeply distrustful of government authority, and therefore provides extensive formal safeguards in the form of complex procedural requirements, and the right of a wide range of citizens to seek judicial remedies if they feel their interests are being disregarded. The ironic result is that laws designed to protect the environment end up making it much more difficult to achieve these ends by delaying or blocking projects needed, for example, to mitigate the effects of climate change. Moreover, the costs of hiring consultants to help in the preparation of EISs, EAs, EIRs, and the like are high, as are the costs of litigation. This inevitably means that wealthier groups and individuals are far better prepared to use the system to protect their interests than are ordinary citizens—and so an other irony: A system ostensibly designed to protect democratic access to decision-making actually magnifies the power of elites.

A Better Way?

As noted earlier, the United States is an outlier among developed democracies in its constitutional structure, and particularly in its extensive use of the judiciary to solve problems that in other democracies are handled by an administrative hierarchy. It is also exceptional in the procedural formality with which it arrives at decisions.

Germany, for example, is a parliamentary democracy with a strong degree of federalism. In terms of the number of veto players, it is located somewhere at a midpoint in the spectrum between Britain’s highly decisive Westminster system that concentrates huge powers in a unitary executive, and the U.S. check-and-balance system. In Germany,

Scoping and review decisions are concentrated in a single authority, which typically has around six months to complete its review. The review itself is incorporated into a broader “administrative act” which grants or denies the project’s application in its entirety, meaning that environmental questions are not eligible for independent court review. To the extent that the administrative act itself can be challenged, courts are largely limited to reviewing issues surrounding the limits of agency authority, not specific conclusions reached by agencies. Furthermore, challenges are limited by relatively narrow standing rules and a one-month statute of limitations, and are seen by expert judges of higher administrative courts who engage in their own fact-finding, ensuring speedy resolution of claims. While the right to intervene via lawsuit is relatively narrow, robust public participation is instead fostered through extensive public involvement in initial project development.¹⁷

This does not mean, of course, that Germany always completes large infrastructure projects on time and under budget. Ordinary politics can interfere with the process there just as in other countries, as in the infamous case of Berlin’s new international airport, which has been interminably delayed. Britain, for its part, has been debating endlessly the construction of a new runway for Heathrow Airport, in which NIMBY concerns have played a major role. Under the civil law that prevails in much of Europe, interest groups can sue administrative agencies and appeal their cases up to the highest civil authority, like the Conseil d’Etat in France. But standing is much narrower and courts are generally much more reluctant to interfere in administrative decisions there than in the United States.

A more detailed comparative study of how other developed democracies deal with infrastructure issues would be helpful in informing the U.S. debate. It is important to note, however, that the U.S. Constitution and political traditions limit the extent to which practices of parliamentary systems like that of Germany can be adapted for use here. Many suggested reforms of the current system involve making it more hierarchical, and reducing the amount of agency seconding-guessing through formal judicial review. All of this is much easier to accomplish in a parliamentary system with a unitary executive that is seen as the instrument of the legislature, which has much more limited judicial review. In the United States, the two branches are often competitive, particularly in an era when they have been held by sharply polarized political parties. Needless to say, both the Left and Right in the United States have a longstanding tradition of distrusting an autonomous executive exercising discretionary power, which is why our cumbersome system of formal checks was created in the first place.

In their 1984 book Implementation, Jeffrey L. Pressman and Aaron Wildavsky described a project in Oakland, California initiated by the Federal government during the late 1960s-early 70s as part of the War on Poverty.¹⁸ The project hoped to create jobs for unemployed minorities in the rebuilding of Oakland’s airport. Despite the relative absence of entrenched actors opposing the project, the government spent several hundred million dollars and the better part of a decade trying to implement it, before finally giving up and pulling the plug. Pressman and Wildavsky note that the reasons for this had to do with the project’s complexity: Several dozen Federal, state, and local agencies took an interest in its outcome, and each required a series of checkoffs that for the most part had to be addressed sequentially. Under the best circumstances, simply obtaining the necessary approvals would have taken four and a half years before the project could have broken ground.¹⁹

In the years since this book was written, the situation, especially in California, has only gotten worse. Most of the checkoffs cited by Pressman and Wildavsky were internal to the executive branch and did not involve judicial review, which constitutes a whole new layer of requirements since added. Interest groups today are also much better organized and can take advantage of a much larger body of case law to manipulate the system.

In addition, our transparency laws have been applied inconsistently, in ways that simultaneously cripple efficiency and yet leave large areas of public participation free of any outside scrutiny. The Federal government is one of the most highly transparent organizations in the world: Under FACA, any agency seeking to consult with an interested stakeholder must have it registered with the GAO, ensure that partisan representation is balanced, and make consultations public and open to the media. This has a deadly impact on agency deliberation, where commissioners must avoid talking to one another in groups for fear of triggering FACA rules. It makes informal consultation with stakeholders vastly more difficult, which is substantively dysfunctional beyond its leading to delays and increased costs.²⁰

None of this public scrutiny applies, however, either to the legislative branch or to the private participants in major infrastructure cases. While legislators hold public hearings when formulating legislation, they control that process and are under no formal obligation to disclose their contacts with interest groups (which often play large roles in drafting the legislation they put forward). And in the formal review process at both a Federal and especially state level, there are no requirements to disclose the identity and interests of those seeking to block legislation. This situation differs from Europe, where there is a much more structured and transparent process for formulating legislation within the executive branch, and much more limited standing for judicial second-guessing.

Concurrently with the launching of a new infrastructure initiative, the next President must therefore seek to rationalize the decision-making process if anything is to be accomplished in the first four years. The United States has too many formal legal requirements for making and implementing decisions, embedded within a system of checks and balances that empowers minority interests to block majorities. Fixing this system will entail (1) increasing hierarchical authority within the executive branch to better coordinate the NEPA review process; (2) limiting the scope of NEPA reviews; (3) using technology to achieve better coordination across agencies; (4) at a state level, at least, reducing standing and focusing stakeholder input on those directly affected by a project; (5) failing that, to increase transparency requirements for parties trying to block or otherwise litigate projects.

Some of these goals can be achieved within the executive branch itself, given sufficient political will. The Obama Administration has in fact sought to streamline the permitting and review process under Executive Order 13604.²¹ Others would require larger statutory changes to accomplish. The politics of statutory change is complex, to say the least, especially under conditions of polarization. Sometimes it takes an exogenous event to shake the system from its low-level equilibrium—perhaps the advent of a new administration brought to power in a landslide can constitute such an event. (Something like this happened in California, when reforms were initiated in the wake of an attempt to block the relocation of the Sacramento Kings to Seattle, Washington.)²²

The second big problem confronting a new infrastructure initiative is how to actually allocate large-scale resources to projects. Transportation bills have classically been places where Congress loves to load up the “Christmas tree” with pet projects reflecting a particular member of Congress’ political clout rather than the project’s underlying economic value. This is the reason that many observers have proposed establishing an infrastructure development bank that would take investment decision-making out of the hands of politicians and delegate it to an impartial board of experts who would decide on the relative merits of different projects. While such an institution could theoretically de-politicize decision-making, delegation to expert groups has not been a popular tactic in U.S. political culture. Given the political system’s degree of polarization, it is hard to imagine Congress coming to agreement on the establishment of such a bank; political disputes over particular projects would probably give way to disputes over appointments to the bank’s board.

There is also possibly a third challenge. The revolution in information technology bears many implications for new efficiencies in infrastructure and, as important perhaps, for integrating infrastructural element in a more efficient system-of-systems. If, for example, electrical grids can be designed to take into account peak and lull periods not just for producing power in homes and businesses, but also for transportation, communications, water and sanitation, and other uses, much greater efficiencies overall can be gained. But unfortunately, no convening platform exists at any level of government to integrate the design, financing, construction, management, and maintenance of smarter integrated infrastructural systems. These systems remain in separate silos with different business models and difference relationships to government according to how they originated and evolved. We can fix and extend legacy infrastructural systems without tackling this problem, and we should do that if we need to. But we would be wasting an enormous opportunity.

Democratic government is of course based on the representation of multiple interests that must be reconciled by the political system. As the U.S. system has evolved in its unplanned way, the accretion of interest groups has indeed had this effect of limiting prospects for strong government. The problem is that it has also had the effect of making much more difficult the achievement of consensus on issues of common concern, like the provision of infrastructure and other basic public goods. As one observer has noted, no individual safeguard built into the current system is objectionable taken by itself; nor does anyone contest aims like environmental protection. The problem is rather the sheer number and complexity of existing safeguards, which leads to huge inefficiencies for the system as a whole.²³ No one would deliberately design anything like the current system if he had to do it again from scratch. Ironically, the vetocracy that was originally intended to limit government now makes it extremely difficult to cut back the system from within, which is the sine qua non of genuine reform.