Big Digs and Soft Budget Constraints
Why do large, government-sponsored projects rarely go according to plan?
Boston’s Big Dig turned the phrase “plagued by cost overruns” into a tired cliché. Initial 1982 estimates put the project at $2.8 billion (about $6.9 billion in today’s dollars). But on completion, the total cost to federal, state, and local governments including future interest payments was a 2016-adjusted $22 billion — over three times the planned outlays. The last Big Dig bond comes due in 2038 — fifty-six years after the project began. The Massachusetts Bay Transportation Authority endures the nation’s highest annual debt servicing bill, with $425 million — nearly a quarter of its 2015 operating budget — going to bond interest. The ultimate victim of this planning disaster is, of course, the taxpayer.
In Massachusetts, the financial burden is passed on to the public through higher gas taxes— a regressive tax on low-income Bay Staters whose bottom lines suffer a higher proportional impact from gas price hikes than do those of higher-income citizens.
It’s not hard to find other projects facing similar plights. Plans for an ongoing extension of Washington, D.C.’s Silver Line have been revised 150 times during construction. These changes, admits the program’s former manager, have contributed to substantial delays and added costs. In Manhattan, a Long Island Rail Road expansion that began in 1998 was expected to cost $3 billion over twelve years. Now, the bill amounts to over $10 billion, and construction won’t take fewer than twenty-four years. And Bertha, the machine boring Seattle’s Highway 99 tunnel, stood still for two years after hitting a steel pipe carelessly left in its path by DOT surveyors.
The Federal government fares no better than state and local authorities at avoiding overruns, and affected projects aren’t limited to roads and tunnels. Construction of the new Department of Homeland Security headquarters is now $1.5 billion over budget and 11 years behind schedule. The Department of Defense spent $5 billion developing and procuring a wear-anywhere Universal Combat Pattern before determining that the redesigned uniform performed poorly everywhere.
The F-35, whose lifetime cost stands at over $1 trillion, has suffered from inadequately defined, unrealistic goals. The Department of Defense did not provide industry partners with clear objectives and has failed to provide sufficient oversight to ensure accountability among contracting firms.
But a lack of sound project analysis does give us one gift: the opportunity for creative headlines. The F-35’s overambitious designers, according to The Atlantic, succeeded in creating “A Weapon that Costs More than Australia.”
The only reliability to be found in large, government-led projects is their consistent underestimation of time and cost. This is not new; by 1825 the Erie Canal had gone 46% over budget, and subsequent expansions would cost over twice what was expected at the time. On-budget and on-time public works projects are exceptions to the norm.
In a world where project cost overruns were randomly distributed around a central value close to estimated cost, those with statistical inclinations would expect about half of all projects to come in under budget and the other half to require additional spending. But in our world, over 90% of publicly run infrastructure ventures exceed cost and time projections.
Oxford’s Bent Flyvbjerg attributes these misapprehensions to “strategic misrepresentation” by public officials eager to secure funding for their own projects. Cost underestimations result from deliberate efforts to secure funding from policymakers who would hesitate to support more involved undertakings.
Other research connects bad estimation with less sinister but nonetheless subpar reasoning. Project experts tend to see their work as unique and neglect what Kahneman and Tversky called “distributional data,” the different outcomes that can result from varying external conditions. Planners therefore fail to look for parallels between their projects and similar undertakings that encountered, for instance, labor, weather, or procurement delays. Another danger is irrational optimism — a planner very much wants to see his project realized and may become deluded as to its feasibility within given constraints. Why should committed exceptionalists expect their programs to any trend but the ideal?
Even if misrepresentation is more inadvertent than strategic, the taxpayer still suffers. A frequent argument during and after over-budget projects is that new roads and defense technologies will improve peoples’ lives and safety — that these undertakings are worth every unforeseen penny that must be spent. But where does the logic of this attempted salvage lead?
Building a better stealth fighter can improve the safety of the American public. Revitalizing aging transportation infrastructure can alleviate urban congestion and mitigate associated economic inefficiency. Surely, a value can be placed on the benefits of these programs. This number is likely finite, so it follows that a maximum viable cost function could be calculated for a given proposal. Governments should seek to maximize the net benefit of their spending, but the very nature of public sector action (or inaction) often precludes optimization. Well-intentioned policymakers should recognize that public sector undertakings are not necessarily the best way to execute such projects. In many cases, shifting some or all responsibility away from bureaucratic administrations can dramatically improve on-time and on-budget performance.
Inasmuch as strategic or inadvertent cost misrepresentation indicts government project planners, it also shows that sponsoring government agencies share the blame. The ballooning bureaucracy that characterizes many public organizations similarly hinders the delivery of on-time, on-budget services and large-scale projects.
The Federal Aviation Administration’s ongoing NextGen program, an effort to modernize the dreadfully archaic American airspace management system, suffers from these ailments. Air traffic controllers in US towers still spend their days tracking aircraft by moving paper strips around on a large board; FAA has only very recently awarded Lockheed Martin a contract to computerize this process for $344 million. Meanwhile, Canada’s non-governmental air navigation service provider began the digital transition nearly two decades ago and now earns over $20 million a year licensing the technology to other nations.
Poor input projections coupled with poorly defined goals have led to a long, expensive slog that scarcely resembles FAA’s initial vision for NextGen. While the administration is authorized to partly self-finance through aircraft user fees, it is forced to continually approach Congress for reauthorization. This uncertainty nurtures a use-it-or-lose-it budget mentality and contributes to administrative inefficiency.
Parkinson, Niskanen, and Kornai
What do American infrastructure administrators and Soviet factory managers have in common? Soft budget constraints. János Kornai, the the economist who first proposed this concept, was once fired from the Hungarian Communist Party’s newspaper after becoming disillusioned with the Party. Entities that enjoy state support make decisions according to an entirely different framework than that employed by their private sector counterparts. Managers in the former class of institutions know that their failures will always be remediated by increased public spending. Why should planners spend time and energy crafting economically efficient projects when they know that the public will ultimately pay any amount?
Government service providers often adhere to Parkinson’s law, propounded in a marvelous satirical piece that The Economist ran in 1955:
IT is a commonplace observation that work expands so as to fill the time available for its completion. Thus, an elderly…
An elderly lady of leisure can spend the entire day in writing and despatching a postcard to her niece at Bognor Regis. An hour will be spent in finding the postcard, another in hunting for spectacles, half-an-hour in a search for the address, an hour and a quarter in composition, and twenty minutes in deciding whether or not to take an umbrella when going to the pillar-box in the next street.
Unconstrained bureaucratic budgets and workforces, he argued, follow a similar pattern and grow unstoppably until hard constraints are imposed.
There need be little or no relationship between the work to be done and the size of the staff to which it may be assigned.
Parkinsonian bureaucratic expansion is multiplicative — when a civil servant feels overworked, he can do one of the following: resign, ask a coworker for help, or transfer work to several subordinates. Parkinson proposes that workers will find the latter most desirable. Resigning could jeopardize the employee’s retirement benefits, and working with a colleague would introduce competition into the market for promotions. He must therefore accumulate more than one helper; delegating to just one would be akin to the promotion-risking scenario.
As the workforce of a given department grows, so does the share of time and energy spent managing and coordinating.
In the 1970s, William Niskanen began to sound the empirical alarm on unconstrained bureaucratic bloat. His budget-maximizing model posits that government administrators seek to increase their Congressional appropriations until reaching a cost-benefit constraint. Niskanen therefore predicts that the equilibrium output of a bureau under this decision-making framework is monotonically higher than that of an efficient, perfectly competitive firm.
While several challenges to Niskanen’s original framework have revealed its deficiencies, the model remains useful in assessing the behavior of many public-sector organizations. There are certainly groups with nobler intentions than budget and workforce expansion, but Niskanen is not wrong in lamenting bureaucratic self-interest. Too often, the government prioritizes self-perpetuation over responsible use of its people’s resources.
What can be done?
Opponents of removing full authority from the hands of the government often cite similar failings in the private sector. Yes, the substantial unforeseen expenses incurred during Boeing’s 787 rollout are an example of a megascale private-sector project gone awry. But incentive-based analysis shows that this should be far from the norm.
From time to time, private sector enterprises face sink-or-swim plights. If a construction firm fails to deliver on building a house, it faces at best a lost customer and at worst an existential lawsuit. But an offending public agency can simply plead for more government funding — only possible to entities enjoying softer budget constraints. Public sector organizations therefore see fewer incentives to employ best-practice risk assessment and market forecasting methods. From this accountability gap ensues a gap in the integrity of estimation.
Compromises like public-private partnerships encourage capitalistic decision making while imposing checks on rent seeking in public goods markets. Uncertainty is distributed between participants; the government sponsor bears responsibility for regulatory obstructions, while the private partner assumes liability for risks associated with the construction process. While unforeseen costs in public works projects are offloaded onto the taxpayer, public-private undertakings shift the burden of payment onto private firms. A profit-maximizing private firm sees incentives to minimize program costs.
For these partnerships to succeed, planners must take care to clearly define desired outcomes. The Federal Chief Information Officer advocates for incremental contracting, but agencies do not always heed this advice. The Department of Transportation Inspector General has criticized FAA NextGen implementation, citing frequent use of large, protracted contracts that do not provide adequate portholes for oversight. By introducing intermediate points of accountability into the development process, project managers can more easily measure progress and mitigate potential failures.
Put concisely by a 2013 McKinsey report,
It is striking to see that — in the absence of private-sector management techniques and private-sector risk takers — public-infrastructure sponsors seldom apply state-of-the-art risk- and project-management tools and techniques, despite the knock-on consequences of being seen to “lose” public money during a time of increasingly constrained public budgets.
Another remedy for cost overruns comes through changing funding structures. Granting partnerships independence from the government appropriations process and allowing self-financing in capital markets can deliver better value and more efficient projects.
Niskanen is fundamentally correct about key elements of the common bureaucratic mindset: indiscriminately throwing money at a problem and hoping that it will go away is rarely the best method. The question policymakers ask the agencies they oversee shouldn’t be “how much more money do you need?” Rather, they must require the public sector to take greater care in where, when, and how they spend their appropriations. The question becomes one of aim, not scope.
Decision makers and planners should pay attention to private sector practices. Since private firms face a much harder budget constraint (bankruptcy), inappropriate spending is curtailed by necessity and survival instinct. In many cases, projects can be delivered more quickly and at lower cost via public-private partnerships or fuller commercialization. Policymakers must realize that they too will benefit from these gains in efficiency; they will enjoy less public outcry over failed timelines and ballooning tax burdens.
Follow @raymondopolis on Twitter.