Version classiqueVersion mobile

The Infrastructure Finance Challenge

Ingo Walter

3. Infrastructure Attributes and Problems of Market Failure

Texte intégral

1Based on the definitions in the preceding section, we consider “infrastructure” to encompass the following specific kinds of facilities:

  • Transport: Roads, bridges, and tunnels; rail systems; airports and air traffic control; harbors and ports;

  • Power and energy: Electrical generation units; high-voltage electrical distribution; refineries and natural and shale gas liquefaction and regasification units; natural gas and petroleum pipelines and distribution centers; and the newer entrants: renewable energy, battery storage, micro grids, and smart transport grids;

  • Water and sewage: Canals and irrigation systems; water purification plants; water pipelines; sewage pipelines; sewage treatment systems; as well as green infrastructure such as wetlands, greening of impervious surfaces for storm-water capture, forest carbon sinks, and watershed protection, among others;

  • Telecom: Landline telephone systems; landline cable and broadband systems; satellite networks; cell/mobile networks; and

  • Social: Public housing; schools; hospitals; prisons.

2Although green infrastructure represents a newer asset class, we nevertheless include it in our definition because of its exponential growth that will be scaled in response to recent global policy commitments as well as by individual governments. China, for example, has developed an aggressive infrastructure plan that includes public- or private-sector financing aimed at tackling its problems with water availability and quality, air pollution and soil degradation.

  • 1 We discuss several important exceptions in the following text.

3As noted, scale economies, particularly those that operate in a concentrated geographical area, make it impossible for a competitive market to provide many infrastructure services. These “market failure” problems — competitive markets’ inability to encompass the characteristics of infrastructure facilities; significant information asymmetries; and significant positive spillover effects — are related to some of these facilities’ typical characteristics, as follows:1

  • Significant economies of scale tend to exist in the facilities’ construction and operation. Larger facilities are less costly on a per-unit-of-output basis than are otherwise similar but smaller facilities.

  • Infrastructure facilities tend to be capital intensive, requiring a large amount of investment relative to their annual output and/or to the amount of labor needed for their operation. This capital intensity usually contributes to their economies of scale.

  • The facilities tend to be long-lived and often last for generations, although many residential and commercial facilities share this characteristic.

    • 2 A local mosquito eradication program is a clear example of a “public good” that meets these criteri (...)

    Most societies perceive infrastructure facilities as serving broad public goals. In some instances they may be true “public goods” for which the facility’s positive externalities are pervasive. That is, additional individuals can receive benefits from the facilities at little or no incremental resource cost and without reducing others’ benefits — and it is difficult or impossible to exclude anyone from receiving those benefits.2 In other instances, the facilities are simply seen as central to the basic functioning of a society. Still, many public goods (e.g., mosquito abatement) are clearly not infrastructure. Indeed, most infrastructure is a rival good, so technically they are “club goods”, not public goods in the strict sense of the term.

4It is worth exploring the connections between these core infrastructure characteristics and the problems they pose for competitive markets. First, however, two counterexamples help illustrate many of the points that need to be taken into account:

5Consider roads and highways. Projects in this area have the characteristics that have just been described, and they are generally considered to be a part of “infrastructure”. But what about the gasoline stations which service the vehicles that use roads and highways? At first glance, these stations are just as essential to transportation and should be considered as part of “infrastructure”. Yet they are generally not included as such. Why not?

6It seems highly likely that their exclusion from the infrastructure category is because gas stations are viable at a relatively small scale; they are not especially capital-intensive; they are not especially long-lived; and they do not have strong “public goods” features. In addition, the direct sale of their output — “pay as you go” — to consumers or end-users is considered a “natural” market arrangement. So the market-driven system can generally develop a near-optimum network of gas stations to accompany a road system, and gas stations thus are not considered to be part of “infrastructure” or part of the “infrastructure problem”. As communication technology advances, the time will soon come when road-use tolling will allow “pay as you go” pricing for road use, and this issue may well disappear.

7A second counterexample is worth considering: Electric power generation — primarily in the form of large-scale fossil fuel, nuclear, or hydro facilities — has traditionally demonstrated the foregoing characteristics and has been considered part of “infrastructure”. A reason for natural monopoly is a minimum efficient scale for a plant that is large compared with the market. But power generation is a service whereby the minimum efficient scale for a plant appears to be decreasing relative to market demand in the immediate area it can service. So in this segment, we increasingly have an opportunity to separate the services provided by the grid from the services of individual generating plants.

8Consequently, a market-based system of power generation units may become feasible. If low prices for natural gas persist, power generation facilities based on natural gas, tidal energy, wind, and solar may proliferate. In response, many of the grid-related infrastructure problems currently associated with electricity generation may diminish in importance.

9What are the key connections between the characteristics of traditional infrastructure facilities and the problems that accompany them?

3.1 Economies of Scale

10Cost efficiency often calls for larger-size infrastructure facilities. Consequently only one or a few facilities may be needed to serve a given market. When prices can be charged for an output, a single supplier of this output means a monopoly seller. Even if a few facilities can serve the market, oligopolistic behavior may interfere with a fully competitive process.

11With monopoly (or coordinated oligopoly) come some important implications. Textbook microeconomics tells us that a monopoly will charge a higher price and sell less output than an otherwise similar competitive group of sellers. Without a competitive process to keep prices reasonably close to costs, society faces a set of choices regarding how the facility will be operated:

  1. Society (collectively) “grits its teeth” and decides to allow a private-sector monopolist to operate unimpeded (unregulated with respect to price);

  2. Society decides to allow a private-sector monopolist to sell the output, but places limits, through some form of regulation, on the price that the monopolist can charge; or

  3. Society decides that the service will be provided by an arm of the state. In this case, a private-sector entity might build the facility under some form of contracting with the government, or the government might build the facility itself.

12Regardless of which route is selected, efficiency issues will arise. For the unregulated private-sector monopoly, a high price will mean that output is lower than it would be if the price were closer to costs. For the regulated private-sector monopoly, setting an appropriate regulated price — so that efficient costs are covered but the firm cannot take advantage of its market power or “gold plate” its costs — is rarely an easy task. For the government-operated facility, the pricing issue is equally relevant, but a challenge arises: How to keep operating costs efficient when the profit motive is not an inherent part of the process?

13These efficiency questions become yet more difficult when a facility is complex and offers multiple services. For example, an airport is an intermediary between airlines and fliers. The airport can charge the airlines for takeoff/landing slots, gate facilities, and hangar/storage/repair facilities. For in-airport services such as parking, food, and travel-related paraphernalia, the airport can either charge fliers directly or charge the vendors that provide these services. Efficiency issues can arise with all of those services and the charges for them.

14Another efficiency consequence of a single supplier is often overlooked but nevertheless important. Unlike in a competitive environment, where sellers offer a variety of qualities and attributes in response to consumers’ differing preferences (e.g., the retail sale of clothing, food, automobiles, etc.), a single provider usually offers a single quality (or at most a limited range of qualities).

15For example, in electricity distribution, different levels of reliability are possible with respect to service interruption (e.g., due to weather-related breaks in above-ground wires). A single provider, however, can provide only a single level of reliability, even if the various consumers within the customer base have differing preferences with respect to reliability levels — and consequently different willingness to pay the costs associated with achieving that reliability.

16This unavoidable limitation on variety generally exists whenever there is a single provider, whether a private-sector entity or an arm of the government, and whether the facility is a true “public good” or perceived as central to the functioning of society.

3.2 Capital-intensity

17An infrastructure facility’s capital intensity describes the size of investments needed relative to output. The presence of economies of scale, for example, concentrates the large investments required for a single facility such as an airport or a single contiguous power grid.

18Regardless of whether the public or private sector controls the facility, debt finance has historically been needed for a large share of such investments. The main issue is that when a government entity retains control of the facility, equity financing cannot include control rights. In turn, the need for debt finance raises a familiar set of asymmetric information problems with respect to financing arrangements. The borrower needs to provide assurances, both before and during the debt contract, that the lender will be repaid. The more difficult it is for the borrower to do so, the more risky the debt contract will appear to the lender, and the more costly this contract will be to the borrower. Under such conditions, the conventional creditor can only lose, never gain, although it is possible to create equity-like securities that bear more risk associated, for example, with fluctuations in demand.

3.3 Long-lived Immobile Assets

19An infrastructure facility’s typical long-lived nature increases the likelihood that its developer will seek long-term financing. But long-term finance itself exacerbates the financing problems just discussed. The longer a financial contract’s term, the greater the possibility that circumstances could change in a way that makes it more difficult for the borrower to service the debt. Consequently long-term financing is generally perceived as more risky and usually is more costly than shorter-term financing.

3.4 Serving General Public Goals

20If an infrastructure facility is perceived as serving broad public goals, then the political process will likely create pressures for low charges, possibly below levels that would cover costs. In addition, to the extent that significant (and likely) economies of scale exist, marginal costs will be below average costs until a capacity constraint is reached, or until the difficulties of managing a large facility kick-in. So, a good economic argument can be made for setting prices that are equal only to marginal costs and that thus do not cover fixed costs.

21In either case, the issue of how to cover total costs becomes relevant and important. A similar question arises in the case of a pure “public good”, for which prices cannot be charged because individuals for political or practical reasons cannot be excluded from enjoying the benefits of the facility’s output.

22In turn, these pricing issues affect financing for infrastructure facilities through creditors’ perceived riskiness of the debt arrangements. This holds true whether the facility is operated by a private-sector entity (and subject to price regulation) or by a public-sector entity whereby the debt-related payments are linked to the cash flows generated by the facility. And for the pure public good, the political willingness of governmental entities to honor their debt obligations raises the same risk issues.

23Consequently, infrastructure finance needs to have an explicit focus on public finance in general, more than the usual idea of a binary separation of tasks between a unified “government” and the market. It must also address the division of tasks between local and national public finance. If in this context there is a role for public borrowing, notably at the local level in federal or otherwise decentralized political systems, what will it take to make this market work better?

24Most infrastructure projects require a physical network that extends through three-dimensional space. Because competition between different physical networks is nearly impossible, the government almost always plays a pivotal role in developing and operating infrastructure. One critical bottleneck that holds up investment in new infrastructure projects is assembling the necessary sites and easements, often by eminent domain. Especially in the case of urban infrastructure, big efficiency gains are available if the government takes the simple step of specifying a grid of public space to be used for trunk infrastructure before private investment takes place.

25Different governments have different capacities and comparative advantages, so the state’s specific role varies from country to country. This heterogeneity means that one important function of the global financial system may be to convert income streams from different countries into standardized securities that can be traded in liquid markets, as discussed below. The need for liquidity complements financial markets’ more familiar functions such as diversifying idiosyncratic risks and resolving asymmetric information problems.

3.5 Sustainability

26Investment in infrastructure has been central to achieving economic development goals, but with it has come a host of environmental and social challenges. For instance, the world’s current on-line infrastructure facilities are responsible for roughly 50% of global greenhouse gas emissions. Infrastructure projects have destroyed ecosystems that protect against flooding (e.g., wetlands) or act as carbon sinks (e.g., forests). They have displaced native peoples, often leaving them in poverty. China’s resource- and infrastructure-heavy growth has made 60% of its groundwater unfit for human consumption, generated air pollution at toxic levels over large swaths of the country and left 19% of its arable land too polluted for agricultural use.3

27The country’s aggressive “take no prisoners” approach to growth was sustainable only as long as the social benefits were seen to exceed the social costs. Even under one-party rule, political sustainability will eventually be called into question once a nation’s material needs have been successfully addressed — unless those costs in the form of deforestation, overfishing, toxic mining spoil, species extinction etc. can be shifted to others. This is trans-frontier environmental degradation through trade and foreign investment. At least domestically, China has developed environmental targets and a public-private partnership investment framework to invest in retrofitting old infrastructure and develop new projects with fewer negative externalities.

28This challenge is not confined to emerging economies, although most developed countries have had the means and the luxury of incorporating long-lasting, high-level sustainability safeguards for more than a half century. Still, neglect and delay along with budget constraints virtually guarantee that old legacies will re-emerge. For example, the 2016 water supply crisis in Flint, Michigan exposed residents to high levels of lead, pointing to underinvestment in environmental and health attributes of basic infrastructure services. Months after the issue arose, no resolution had been found regarding accountability or remediation responsibility.

29On a global level, the UN Sustainable Development Goals (SDGs) include criteria on clean energy, infrastructure, sustainable cities, and climate action that will require a new 21st century approach to infrastructure. This new approach covers broad objectives, such as reducing carbon and other environmental impacts, as well as support for new technologies such as electric self-driving vehicles and transportation grids in urban centers.

30“Green” costs money, however. In the years following the 2007–2009 financial crisis and the Great Recession, the need for project-related financing grew steadily. According to Brookings [2015], the incremental costs to “green” these investments and make them sustainable could be as much as $4 trillion in gross terms, not including operational savings and positive externalities. Perhaps in the end, the benefits significantly exceed the costs. But it is the costs that must be credibly defined and financed.

3.6 The Broader Implications

31The brief overview contained in this section on the economics of infrastructure development offers some further implications that can be summarized as follows:

  • Getting infrastructure “right” is a difficult analytical challenge because of issues about “scale” relative to market size.

  • Government will always be involved in infrastructure projects. Because the capacity for governance and the sophistication of the supporting political process will vary across jurisdictions, different systems for providing infrastructure will be appropriate in different contexts.

  • Sustainable and green infrastructure aimed at solving environmental and social challenges is a growing priority.

  • Technological innovation is opening up many new options.

  • Some types of service, such as power generation, that were previously infrastructure services may increasingly be provided in a competitive market.

  • Demand exists for new types of infrastructure, such as a grid of fiber optic lines.

  • New information technology could allow much more detailed monitoring of use as well as new categories of user fees based on these measures, which can address some of the free-rider problems in certain types of infrastructure.

32Infrastructure financing issues in light of generic drivers are as real and substantial as the projects themselves. Although understanding the nature of these problems does not automatically eliminate them, improved understanding can point toward design and implementation strategies that may lessen their severity. The world’s infrastructure needs are of such magnitude that even modest improvements in financial efficiency could create significant gains.


1 We discuss several important exceptions in the following text.

2 A local mosquito eradication program is a clear example of a “public good” that meets these criteria; national defense is another frequently cited example.

3 Goldman Sachs Group, “China’s Environment” July 15, 2015 at

Rechercher dans OpenEdition Search

Vous allez être redirigé vers OpenEdition Search