Given the enormous costs of infrastructure investment, careful selection of projects is critically important. Three chapters from the book Infrastructure Economics and Policy: International Perspectives published by the Lincoln Institute explain how cost-benefit analysis has become a nearly universal evaluation tool adopted by governments and international financial institutions, and how investment decisions are shaped by cognitive biases and political factors.
The Evolution of Cost-Benefit Analysis
Don Pickrell, Chief Economist of the Volpe National Transportation Systems Center under the U.S. Department of Transportation, introduces the evolution and methodological challenges of cost-benefit analysis in the chapter on economic appraisal. The French engineer Jules Dupuit in the mid-19th century is widely recognized as the originator of this analytical approach. His core research question was how to judge whether investment in new infrastructure such as a bridge was worthwhile. He proposed that the standard of judgment lies in whether the benefits accruing to direct users of the bridge — including savings on time, labor, fuel and other expenses — exceed the costs of construction and maintenance.
For a long time, certain scholars have criticized this exclusive focus on direct user benefits for overlooking the wider economic impacts generated by infrastructure. Advocates of the broader impact perspective argue that time savings and other gains enjoyed by direct users can be passed down to downstream enterprises, enabling further productivity improvements. Ignoring such ripple effects will lead to systematic underestimation of a project’s overall social value. Some researchers have constructed economy-wide mathematical models to quantify these wider economic impacts, incorporating estimates of how shifts in infrastructure efficiency affect productivity across all industrial sectors. Nevertheless, this model-based method for calculating both direct and extended economic impacts failed to gain popularity among researchers and governments until a decade ago. Pickrell speculates that the requirement to estimate or assume a large number of parameters hindered its popularization; such macroeconomic models are also less suitable for civil servants tasked with choosing among discrete infrastructure projects, and are better applied to evaluate the outcomes of economy-wide reforms.
Skeptics of this framework hold that so-called wider economic impacts are often not incremental new gains, but merely transfers of value from primary end users to other parties. For instance, the opening of a new bridge typically pushes up local land prices and stimulates new residential and commercial development nearby. Rising land values stem entirely from reduced travel time brought by the bridge; counting both travel time savings for users and land appreciation as separate benefits amounts to double-counting the same value gain.
A consensus has taken shape in practical application: wider economic impacts may be included in benefit calculations without double-counting if they correct pre-existing inefficiencies in affected downstream markets. Typical examples include environmental costs and benefits imposed on third parties, general productivity growth driven by urban agglomeration, and price reductions resulting from heightened competition after monopolies are broken up.
Cost-benefit analysis has evolved into a standard evaluation tool for major public infrastructure projects used by governments and international financial institutions worldwide, which represents remarkable progress. Its widespread adoption is largely attributable to continuous methodological refinements over the past century. Despite its utility, few academic studies have examined the real influence of cost-benefit analysis on governmental project selection. Limited existing research yields pessimistic results: top-ranked projects are rarely chosen for implementation. Defenders of cost-benefit analysis, however, maintain that its core strength does not lie in picking the single optimal option over the second-best alternative, but in eliminating the worst-performing projects from the candidate list.
Optimism Bias
Cost-benefit analysis is far from flawless. In Chapter 7, Oxford University Professor Bent Flyvbjerg and statistician Dirk Bester present extensive statistical evidence demonstrating consistent systematic biases in appraisal practice: costs are drastically underestimated, while projected benefits are severely overstated. Their impressive dataset covers 2,062 infrastructure projects of six investment categories across 104 developed and developing countries, all commissioned between 1927 and 2011.
Forecast inaccuracies are frequently attributed to short-term changes in project design or external environments, such as unplanned expansion of project scope, higher-than-expected inflation, weaker market competition than anticipated, and similar variables. Flyvbjerg and Bester, however, argue that the root causes are well-documented behavioral limitations, particularly optimism bias and overconfidence bias. This finding implies that improving forecasting accuracy will prove difficult, as such cognitive flaws are deeply embedded in human nature. The pair put forward concrete reforms to cost-benefit appraisal, including formal debiasing procedures and introducing financial incentives and penalties tied to the forecasting accuracy of consulting analysts.
Political Drivers of Overspending or Underinvestment in Infrastructure
In Chapter 8, John Donahue of Harvard Kennedy School draws on public choice economics to explain the political mechanisms behind excessive or insufficient public infrastructure expenditure. Public choice theory posits that individuals rationally pursue self-interest within democratic societies governed by various voting systems. Therefore, whether public choice processes lead to overspending or underinvestment hinges on two variables: how well-informed voters are, and the design of voting rules.
For example, Anthony Downs, a prominent public choice scholar, argues that public spending tends to fall below optimal levels because voters consistently underestimate benefits to a greater degree than they underestimate costs. Mancur Olson holds that public goods expenditure is driven by political coalitions, which operate most effectively when they remain small or can offer exclusive perks to attract new members — roadside assistance from the American Automobile Association serves as a typical example. Gordon Tullock rigorously illustrates that majority voting for parallel infrastructure projects creates incentives for logrolling, where proponents exchange mutual support, ultimately resulting in excessive public expenditure.
Collectively, these three chapters elaborate the strengths and limitations of cost-benefit analysis, the prevalence of human-nature-driven optimism bias, and how public choice dynamics shape the scale of public infrastructure spending. Deeper understanding of these factors can substantially improve public infrastructure investment decision-making.
About the Authors
José A. Gómez-Ibáñez, Derek C. Bok Professor Emeritus of Urban Planning and Public Policy, Harvard University
Zhi Liu, Senior Fellow and Director of the China Program, Lincoln Institute of Land Policy
They serve as co-editors of Infrastructure Economics and Policy: International Perspectives.
