5. Implications for Selecting Policy Instruments
5.1 Matching Possible Instruments to Particular Issues
This section articulates a set of principles as a guide to infrastructure policy development, taking account of both growth and sustainability objectives. It aims to provide a set of practical guidelines that can inform both those involved in policy in central and local government, and also show those in the private sector where government priorities are likely to lie.
It also draws out the implications from the previous sections to identify what information will be needed to monitor progress against overall objectives, and how to interpret it in a policy relevant way. It will inform the Infrastructure Audit about which pieces of information identified in the audit is of most significance for policy interpretation, and also identify types of information that may not currently be available, but which might feasibly be collected.
5.1.1 Principles
The following section outlines a set of "principles" or considerations to be taken into account when designing infrastructure policy. It is not a blueprint for policy so much as a set of marker posts for defining what policy can do.
The underlying principles of this infrastructure policy framework are:
- Sustainable development is about improving the outcomes from use and configuration of available resources (physical capital, natural capital, social capital, cultural capital), and as such is fundamentally an economic issue. Economic analytical frameworks, particularly with respect to institutional analysis, externalities, and operations of networks, are appropriate to guide infrastructure policy, with some modification to allow for wider issues than those commonly regarded as affecting the economy.
- Where they operate well, markets provide the best means of allocating resources so that they move to their most valuable uses, but market failures and externalities limit their achievements in practice. Improving markets by remedying market failures and externalities is a primary means of improving infrastructure's contribution to welfare, and is appropriate for both economic efficiency and sustainable development.
- Policy needs to focus on the most important externalities that can be affected by policy, rather than on the total level of resource use or impacts. In the case of infrastructure failures like congestion, this means focusing on the marginal impacts of additional congestion, and on the marginal benefit of congestion reduction, rather than the aggregate impact of congestion as such.
- Policy needs to support a macro-economic climate conducive to effective resource allocation and use, and avoid undermining it by inducing expenditures that exacerbate business cycles and inflationary pressures on input resources.
- Policy needs to be flexible enough to allow different responses according to varying conditions in different sectors and regions, rather than imposing a one-size-fits-all approach that will lead to varying incentives and compliance costs around the country.
- In distributing risks from infrastructure development or non-development, policy should strike a balance between placing risk on those who have most influence on the outcome (so as to avoid moral hazard) and placing it on bodies who are best placed to spread and bear the risk (so as to avoid the opportunity cost of excessive precaution).
- Similar principles apply in distributing responsibilities across those involved in supplying services through infrastructure. Issues that primarily concern internalised effects can be left to private entities (e.g. investment and operational decisions). Effects that demand a collective policy response can be devolved to the lowest level of collective authority capable of dealing with it (the principle of subsidiarity), with the proviso that localised authorities face tighter constraints on the resources and funding they can bring to bear.
- Infrastructure policy is only one part of broader socio-economic policies, and depends on other policies articulating the settings within which it can operate.
In practice, infrastructure policy requires a demand orientation in both evaluation and operation of infrastructure investments, with performance indicators that reflect the quality of service and user satisfaction, not just measures of physical assets and financial performance. Sustainability also requires attention to changes in long term wealth stocks rather than just annual income flows. This implies a sustainability reflective policy framework needs information that allows monitoring and analysis of:
- Broad trends in population, economic activity and other influences on the demand for services currently provided by infrastructure, including both predictable trends and the likelihood of shocks to the supply system. This establishes an expected business as usual setting against which to compare trends in infrastructure supply and susceptibility to change.
- Infrastructure capital stock, not just its current state and quality but also its rates of depreciation, renewal and upgrade, and the prospects for future prices, to explicitly show long term sustainability of infrastructure services.
- Supplementary indicators of externality effects that fall on the natural environment or on society generally, to explicitly take account of environmental, social and cultural impacts and assist in keeping them within prescribed limits of acceptability for sustainability.
- Macro- and micro-policies and their role in ensuring an open and competitive economy that allows efficient resource allocation, specialisation and firm contractual agreements, so as to encourage free association and partnerships in economic activity.
5.2 Determining Cross-Sectoral or Sector Specific Responses
While, as we have argued throughout this report, there are high level similarities that mean that policy for infrastructure will have a common shape, sectoral differences are also important. So, determining when to respond in terms of the sector and when to be more generic is a component of any practical infrastructural policy.
In principle, a series of tests can be drawn up. These relate the effects of the various alternatives to the key objectives of the policy. Thus, the overall criterion would be:
Would a tailored (sector specific) intervention of this type be better from the national viewpoint (assessed against the four aspects of interest) than a generic response?
So, when the question relates, say, to general matters of regulatory control, or social accessibility, an intervention that covers all infrastructure will probably be appropriate as a cross-sectoral response is likely to better serve the public interest by providing consistent weightings to individual outcomes from resource allocations. On the other hand, where the question seems to relate more to particular attributes of the sector, like the electromagnetic effects possibly related to cell-phone towers, a specific intervention that relates to the sector might be more effective.
5.3 Determining Who is Best Placed to Achieve Change
5.3.1 General Considerations
The organisations best placed to achieve particular changes in outcomes depend on a range of different capabilities. These include:
- Managerial skill - rapid and flexible decision-making, due to clear structure of incentives;
- Productive efficiency - lower production and delivery costs stemming from appropriate incentives (such as the profit motive in commercial operations);
- Dynamic efficiency - the motivation to invest in and maintain capital equipment needed to expand and introduce technological improvements;
- Accountability to customers - the motivation and flexibility to adapt to changing markets and improve service quality, which is usually enhanced in competitive markets.
- Financial autonomy - freedom from fiscal risk for government.
In the context of New Zealand infrastructure services, the range of organisations that could be involved with infrastructure policy and provision include:
- Government department
- State owned enterprise
- Public utility
- Local government business unit
- Local government council controlled organisation
- Private contractor to public operator
- Private enterprise
The choice between these organisations depends partly on the nature of effects being managed - e.g. those of local, regional or national significance, as discussed in section 3.4 - and partly on the financial circumstances of the governing body itself. For instance, provision of infrastructure through state owned enterprises or local government business unit is generally intended to allow a degree of government control over, and benefit from, the infrastructure, while creating more incentive for commercial efficiencies than would be obtained in a department of the government itself. Using private contractors or franchise operators over public assets seeks to further the realisation of commercial efficiencies, but its effectiveness depends on the ability to tightly define the objectives and outputs sought from the operation to those where the contractor can expect to appropriate a return on its efforts.
5.3.2 Implications of Risk for Assigning Responsibilities
The overriding principle for risk sharing involves striking a balance between those who have most influence over the risky outcome, and those who can bear risk most easily. These characteristics do not generally reside in the same people. Operators of infrastructure facilities will often be best placed to influence achieved returns through their performance and it is appropriate that they bear some of the risk, to avoid the moral hazard that they will operate in ways that create costs for others when risk is borne elsewhere. However, an operator with substantial stake tied up in the facilities themselves may be less well placed to bear risk than a government that can spread it over a wide body of taxpayers, each of whom has only a small proportion of their wealth dependent on the facilities at risk.
The exact mix of risk sharing between government and operator varies with the circumstances of particular facilities. With telecommunications networks, for instance, where services are excludable and subject to market supply, a network operator has strong influence over business success and sustaining demand, by maintaining the quality and range of services. In such cases it could be inefficient and counterproductive to quality for government to assume the risk of changes in future demands. In the case of toll roads, however, an operator has less influence over demand, which is affected, among other influences, by decisions that government may make on provision of the surrounding road network and the level of public transport subsidy. In public-private partnerships for roads it has been common practice for governments to guarantee a level of revenue, should demand fall below initial expectations.
Allocating the risk to government can improve outcomes if government is aware of, and responds to, the financial incentive created by contingent liabilities, but government bodies are widely regarded as less responsive to such incentives than private organisations. Governments may also have less incentive to seek opportunities to reduce risk through diversifying and hedging. Neither can it be presumed that government is necessarily best placed for bearing risk. Large firms with many shareholders, each of whom has a relatively small individual stake in the particular infrastructure operation, may be at least as able to absorb the risk as a government. This is even more so when there is substantial foreign shareholder involvement, because foreign shareholder wealth will have even less exposure to, and correlation with, the vagaries of local market conditions.
While the optimal risk spreading will vary with the detailed characteristics of different infrastructure services, the above principles point to some likely directions:
- Demand risk (that revenues are less than expected) is often better borne by the facility operator than government, where service attraction is under control of the operator. If government assumes this risk, it also exposes itself to the risk that a downturn in demand will increase its liability at the same time as it faces reduced tax receipts and increased demands for other spending (e.g. social welfare supports).
- Political and regulatory risk (that future government actions subsequently undermine the return from the facility) are more under the control of government than of the operator. While the investment climate would be improved if government avoided creating wealth loss and uncertainty for investors, it is difficult for government to credibly commit itself against future policy changes that might affect an investment. To do so would also imply a pre-eminence for the investment over other activities in the community that have no such guarantee against policy impacts.
- Quasi-commercial risk (that a state-owned supplier or purchaser may default on contracts with the investor under political influence) may provide a justification for government to guarantee returns.
- Demand and construction cost guarantees, although commonly sought from government, have weak justification in terms of this risk sharing rationale. A developer or operator usually has more control over construction costs than government, and government is only one of many influences over future demand, and not necessarily the dominant one.
Although offering guarantees for infrastructure investment provides short term fiscal relief for government, it creates a long term potential liability that may not be fully appreciated by decision makers. Cash subsidies or similar financial instruments (tax relief etc) are generally more transparent and easier to target. Changing the form of guarantee may also shift potential liabilities between government and operators. For instance, altering the length of an operating concession according to actual returns experienced (e.g. a shorter length with higher than expected demand, longer length with lower than expected demand) provides a way of ensuring returns to operators without creating large contingent liabilities for government.
An implicit precept of sustainable development is that, without regard to sustainability principles, risks will be shifted away from those who control their creation, either by creating liabilities for future generations to deal with, or by generating effects beyond the boundaries of market transactions, on the environment, on social cohesion and on cultural integrity. Risk spreading has the characteristics of other externalities, and this provides guidance as to how such risk can best be shared.
For risks that affect financial returns on investment, an optimal distribution of risk resembles the optimal allocation of goods and services to private or public good provision. For services with private good characteristics that are delivered through markets and mostly commercially priced (e.g. telecommunications, ports and airports), financial risks are most appropriately managed by the facility operators themselves. For services that are more in the nature of common pool facilities or public goods, there is a stronger case for government to assume some of the risk, but certainly less than all the risk if moral hazard on the part of the operator is to be avoided.
Infrastructure is of such a scale that most providers are large companies or state owned enterprises able to spread their risks over a wide pool of shareholders or taxpayers. Compared to these entities, local government bodies are not well placed to absorb risk, particularly where insurance markets are incomplete, and there is little scope for local bodies to be assuming financial risks of infrastructure development. However, while financial risk may often be diversifiable more effectively by private bodies than by the state, governments both local and central will often be seen as an ultimate guarantor of supply in their jurisdictions, and they would find it difficult to absolve themselves entirely from responsibility.
Environmental risks can be shifted back to infrastructure providers through such means as performance bonds and liability rules that increase incentives to avoid environmental damage. When social and cultural outcomes are less than expected, however, the impacts can only be borne by the communities affected, and the risk resides with the communities of interest.
Back to Top