33482 Part III. Special Topic: EU8 v 3 PPPs--Fiscal Risks and Institutions12 Public­private partnerships (PPPs) are popular around the world, in part because they allow governments to secure much-needed investment in public services without immediately having to raise taxes or borrow. In the EU8, they can help governments undertake investment without breaching the Maastricht criteria for debts and deficits. Yet PPPs pose a fiscal danger as they do create fiscal obligations. The danger of PPPs is that governments' desire to avoid reporting immediate liabilities may blind them to the future fiscal costs and risks, which increases fiscal vulnerability. The extent of the danger depends on the fiscal institutions that shape and constrain government decisions toward PPPs - that is on factors such as fiscal targets (including the Maastricht criteria), budgeting procedures, accounting and auditing standards, and the assignment of responsibilities for fiscal decisions among different parts of government. Better fiscal institutions can increase the chance that PPPs will be well designed and appropriately used--at the end of the day meaning real fiscal savings and higher growth. Other than Eurostat's guidelines in this respect, there are currently no tools to impose prudent treatment of PPPs legally, and the imposition of transparency procedures and discipline for PPPs has to come from the side of the governments on a voluntary basis. 1. Introduction There has been a renewed interest in the use of public­private partnerships to improve economic efficiency and to satisfy large investment requirements in the new EU member countries from Central Europe and the Baltic countries. This interest was perhaps stimulated by attempts to address the difficult challenge of reducing the existing disparities in the quality and availability of public services between old and new members of the EU while maintaining fiscal stability or, in some countries, conducting fiscal adjustment. At the start of their transition to market economies, EU8 countries, like other countries from the region, inherited extensive infrastructure networks that were in serious disrepair and ill-suited to their transforming economies.3 Since the early 1990s, investment in infrastructure in EU8 countries plus Romania and Bulgaria was estimated at 100 billion by the Berlin- based DIW economic research institute. Public investment in 2003 was close to or above 3 percent of GDP in all EU8 countries, except in Latvia where it was 1.5 percent of GDP. Recognizing increasing investment needs, public investment as a share of GDP is projected to increase to 3.5 percent on average for the EU8 in 2006. While the average public investment level in the EU8 is slightly above the level in EU-15 (around 2 percent of GDP), investments required to bring the EU8 infrastructure standards to the average for the union entail much larger costs. The long-term needs for transport infrastructure alone in the EU8 countries and two other accession countries are estimated at nearly 100 billion.4 Furthermore, according to estimates by CASE (2005), the environmental investment needs of the EU8 are estimated at 47­69 billion (Poland 22­45 billion, Hungary 10 billion, and the Czech Republic 9.4 billion). Further increases in public investment may be constrained, as meeting the Maastricht criteria and a successful experience within the Euro area require fiscal prudence. In 2003 and 2004, four of the EU8 countries (the Czech Republic, Hungary, Poland, and Slovakia) had budget deficits higher than 3 percent of GDP, the euro-area benchmark. At the same time, the three Baltic countries have strong fiscal positions reflected in small deficits or overall surpluses. Public debt as a proportion of GDP was lower 1PPPs are defined here as privately financed infrastructure projects in which a private firm either: (i) sells its services to the government; or (ii) sells it services to third parties with significant fiscal support in the form of guarantees. 2Based on a draft chapter by Hana Brixi, Nina Budina, and Timothy Irwin (World Bank)for a report on fiscal reforms in the EU8. 3EBRD (2004). 4EC (2001). 1 than the 60 percent benchmark in all the EU8 in 2004--in the Baltic countries, it was about 14 percent, while in the rest of the EU8 it was a little over 40 percent. Most of the EU8 are continuing the process of fiscal reform, in some countries driven by concerns about medium-term fiscal sustainability and the need for fiscal consolidation, in others by concerns about the quality of fiscal policy. While the EU8 do not yet participate in the single currency, they are required to fulfill convergence criteria, including relating to the sustainability of the government financial position, in order to qualify for the adoption of the Euro. Given the relatively large initial imbalances in the Czech Republic, Hungary, and Poland, these three countries have to undertake the largest fiscal adjustments to get below the threshold of 3 percent of GDP by 2007/08. Nevertheless, general government deficits in excess of 3 percent of GDP are projected to remain in 2006 in all three countries. Given the vast infrastructure needs in all EU8, the issue of how to create fiscal space for additional public infrastructure spending is becoming critical. Heller (2005) defines fiscal space as "the availability of budgetary room that allows a government to provide resources for a desired purpose without any prejudice to the sustainability of the government's financial position." Creating fiscal space is critical when additional fiscal spending would likely boost medium-term growth and eventually pay for itself in terms of future fiscal revenue. The fiscal space for additional infrastructure investments in the EU 8 will be constrained not only by the need in several countries for fiscal consolidation, but also by demographic pressures on social spending, the need to increase resources for higher education and knowledge economy development, and the desirability of lowering labor taxes to stimulate employment. At the same time, there is scope to free resources through curtailing remaining subsidies, privatization (which would lower future public investment needs) and better prioritization of the public investment program, public administration reform, and last but not least, rationalizing social assistance programs (while pension reforms can be associated with cost in the short-medium term to the extent funded pillars are introduced). While some EU transfers will also help create fiscal space (notably cohesion funds, and to some extent preaccession funds), structural funds are subject to the additionality and co-financing requirements and thus a drag on the budget. There have been proposals to adjust the SGP to target the current account balance of the government, which allows borrowing to finance public investment.5 6While there are arguments in favor of such an adjustment (the creation of productive assets would generate higher output and taxes while possibly contributing user charges over the long run, with the full impact of public infrastructure spending on long-run solvency only possible to assess by the changes in the public sector's net worth), allowing borrowing for public investment may be problematic if it overlooks the quality of public investment, possible short-term financing constraints, or conflicts with the need to rein in excess demand. Further, it may also distract attention from difficult fiscal reforms related to reprioritization of expenditure, including cuts in non-priority current expenditure, dealing with fiscal risks and contingent liabilities related to structural reforms and PPPs and eliminating off-budget expenditure, including on infrastructure. Freeing public investment from any fiscal constraint may also invite creative accounting, or classifying current spending as investment, and thereby excluding it from fiscal targets. 5See Blanchard and Giavazzi (2003), Buiter and Grafe (2002) and Serven (2005). 6Note that such a change to the fiscal targets is also not a near-term option, as this would require a constitutional change. 2 In what follows, we focus on whether and when using public private partnerships (PPPs) can create real fiscal space for additional infrastructure investments in EU8. 2. PPPs in the EU8 and their fiscal effects PPPs are not new to the EU8. Indeed, the very first railway in Europe with a government-guaranteed return ran from Warsaw to Vienna.7 But, as in other countries, PPPs have returned to the limelight in the EU8 in the last few years. For example, Hungary has entered into long-term purchase contracts to secure a prison and several university buildings, at an estimated total fiscal cost in present value terms of 483 million (Hungary Budget Act 2005). Hungary has also entered into a long-term purchase contract under which it will make availability payments to the M5 motorway company, conditional upon the company providing an operational motorway meeting specified performance criteria.8 The estimated present value of the payments is 914 million. The two amounts sum to about 2 percent of GDP. Both Poland and Hungary have given guarantees to toll roads. The first privately financed toll road in Poland (at least in the modern era) is a segment of the A2, a motorway under construction running from the German border to Warsaw and beyond (estimated project costs are about 745 million). Hungary originally provided a guarantee to the M5 motorway (which now benefits instead from availability payments). The government agreed that if toll revenue fell below a specified threshold for any reason, it would extend to the concessionaire a loan that would not have to be repaid before the concessionaire's senior lenders were repaid in full. When the road opened, revenue was below the threshold and the government had to make payments, which it was later able to reduce by discouraging traffic on alternative public roads.9 In other cases, EU8 governments have sought private investments in infrastructure that have not required guarantees or long-term purchase commitments. At least one such project nonetheless had significant fiscal implications. The M1/M15 motorway in Hungary was financed purely on the back of forecast toll revenues, but eventually the government took over the road and assumed part of the debt (about a quarter of a percent of GDP). Lastly, public enterprises in the region have entered into long-term purchase agreements that are like PPPs even if they are not usually described as such. These agreements create contingent obligations for the government. In many countries, governments guarantee that the private partner will receive the payments it is due from the public enterprise. Even when governments do not give formal guarantees, however, they may feel obliged to honor the commitments of public enterprises. PPPs can have both real effects on a government's fiscal position and illusory effects that reflect only the nature of government's fiscal reporting. Disentangling the two is not always easy. A government that uses public finance to build infrastructure needs to make large upfront payments for construction. Unless the government increases taxes, its reported deficit and debt immediately increase. A government that uses a PPP, on the other hand, usually need not pay anything upfront, which makes this option very attractive. But fiscal comparisons between public finance and PPPs can be deceptive because: · the true, long-run fiscal effect of publicly financed public services is not always reflected in measures of government's fiscal position and fiscal performance. · even though PPPs do not increase the government's debt they do create direct or contingent fiscal obligations. Even when governments make no explicit financial commitments, they can feel obliged to bail out private infrastructure firms that become financially distressed. A useful way of thinking about fiscal liabilities associated with PPPs is to group them by whether they are direct or contingent and whether they are explicit or implicit (Table 1--Annex). 7Westwood (1964). 8See http://www.ebrd.com/enviro/eias/hungary/35264.htm. 9EC (2004). 3 It is clear that PPPs can have fiscal costs. In some cases, the fiscal costs of PPPs seem to be about the same as the costs of publicly financed investment; it is just that the liabilities are off balance sheet. But the substitution of PPPs for publicly financed investment can have real fiscal effects, either positive or negative. First, they can change the government's net worth (the difference in present values between its expected revenues and its expected expenditures). Second, they can change the government's indebtedness, holding net worth constant. Net worth, which determines whether a government is solvent, is probably the most fundamental fiscal consideration. But indebtedness increases a government's financial vulnerability, heightening the risk of adverse changes in net worth, and matters especially to governments that are already highly indebted. When the government's liabilities in a PPP are off-balance-sheet, and the government is constrained by fiscal rules such as the Maastricht criteria, a PPP can allow the government to undertake more investment than it otherwise could. In this case, the PPP's effect on government net worth depends on the circumstances. If the project generates no user fees for the government, it could increase the government's net worth only if it stimulated economic growth so much that the present value of increased tax revenues surpassed the present value of the government's payments. If the government used the asset procured with the PPP (such as a road on which the government made availability payments to the private partner) to generate user fees, the PPP is more likely to have a positive effect on government net worth. Second, PPPs in which the government is the purchaser of the output (under a long-term contract) can increase the government's net worth if they reduce the total cost of providing the service. Private companies are often better than governments at building infrastructure assets on time and on budget and at operating and maintaining them in a low-cost and efficient manner. That creates an argument for allocating the risks of operations, maintenance, and construction to a private firm, so that the private firm's finances, not the government's, depends on the costs of operations, maintenance, and construction. When a private firm bears these risks, project costs should be lower, and this should reduce the amount the government has to pay. Third, PPPs may increase the government's net worth if they make it easier for the government to raise or introduce user fees. Tolls and tariffs often do not cover the full cost of water, power, and transport infrastructure, and efforts to raise user fees toward cost recovery often fail because of the importance of the services and the fact that prices can be kept below costs without causing the immediate cessation of services.10 When services are publicly financed, governments often plan to raise user fees, but find it hard to carry out the plans in the face of political pressures. Lastly, it is worth noting that PPPs can reduce a government's (real) indebtedness, even when they leave its net worth unchanged. Even if the government gives some guarantees, its total liabilities, along with its assets, will be lower under a PPP than under public financing. This will not matter much to governments with low debt, but could be important to others. Distinguishing between the real and apparent fiscal effects of private financing requires two things. First, it requires governments to estimate the true fiscal consequence of public financing, taking into account future revenues as well as current investment costs. Second, it requires governments to estimate the cost in present values of their commitments in PPP contracts. This is relatively straightforward for long-term purchase contracts in which the government's payments are fixed in advance. The simplest option is to assume that the private partner will achieve the performance targets it is set. All that needs to be done, then, is to determine the payments required by the contract and calculate their present value at the risk-free interest rate appropriate to the term and currency of the obligation. If it was reasonable to suppose that the private partner would not fully meet the performance requirements, the government could estimate the expected penalties that could be deducted from the availability payments. Estimating the cost in present value terms of guarantees and 10For a discussion of this problem see World Bank (2004). 4 other forms of contingent support is more difficult, but often possible. Option-pricing formulas can be used to estimate the cost of simple guarantees. 3. How fiscal institutions affect the fiscal cost of PPPs Fiscal institutions affect the use, design and, ultimately, fiscal cost of PPPs. Fiscal institutions are conventionally defined as institutional arrangements and management practices that relate to public resource allocation, resource use, and financial management. The nature of these arrangements and practices directly affects government incentives, information, and capacity. Incentives As EU8 countries seek to generate fiscal savings while promoting investment, their fiscal institutions affect their incentives toward the use and design of PPPs. Conventional fiscal institutions tend to promote incentives to: · Favor PPPs even when public investment would deliver equal results at a lower cost in the long term. · Accept risks (that is, offer explicit and implicit guarantees and long-term purchase contracts) rather than providing cash subsidies under PPPs.11 · In the design of PPPs, let the public sector accept risks that the private sector is more suited to bear. This is largely because conventional fiscal institutions devote weaker scrutiny to non-cash fiscal support and long-term obligations compared to cash-based support and immediate outlays. The difference mainly relates to the rules for fiscal reporting, accounting, and budgeting, for measurement of fiscal savings, and for government accountability with respect to fiscal performance. Measuring fiscal savings (and fiscal adjustment) in terms of immediate impacts on the deficit and debt that are not adjusted for government risk exposures has been known to encourage governments to provide contingent forms of fiscal support and assume risk and long-term obligations in exchange for short-term reductions of cash spending (Irwin et al, 1997; and Brixi and Schick, 2002). Hence, with respect to infrastructure, measuring fiscal savings in terms of conventional government deficit and debt (that is without considering the future fiscal cost of contingent liabilities, such as guarantees, and long-term obligations, such as take-or-pay contracts) creates an illusion of fiscal savings when investment and services are delivered without immediately raising the budget deficit and government debt. This illusion makes government risk exposure and long-term obligations under PPPs look "cheap" compared to public financing and cash subsidies. Since the fiscal cost of PPPs typically surfaces in the longer term, the illusion holds even when countries develop a medium-term fiscal framework. Although not explicitly captured by ESA95, some guarantees in EU8 countries are included in fiscal accounts under the Maastricht criteria as countries have expanded their definition of general government. Specifically, by expanding the definition of general government, these countries have brought the cost of quasi-fiscal operations implemented through extra-budgetary funds and off-budget agencies into their general government deficit and debt figures under the Maastricht fiscal framework. The commitment to bring extra-budgetary funds and off-budget agencies of a fiscal nature within the scope of general government has been, however, uneven across the EU8 countries. The Czech Republic and Slovakia may be considered the leaders in this regard. Commendably, the Czech Republic since 1997, after publicly disclosing its large contingent liabilities arising from the so-called transformation institutions,12 off-budget funds and credit guarantees, has brought these into the Maastricht fiscal framework. Transformation institutions and off-budget funds 11As the previous section illustrated, private investors often ask the government to provide financial incentives in the form of cash subsidies and/or risk sharing as a condition for their involvement in infrastructure. 12Transformation institutions had been created as off-budget agencies to borrow, issue guarantees and finance government support programs for banks, enterprises, and other entities. Some of transformation agencies were covered by an explicit government guarantee, others were considered backed by the government implicitly. 5 have been either dismantled or scheduled for dismantling. Those that have remained in operation are now accounted for as part of general government, and the cost of their activities thus directly affects the deficit and debt figures reported under the Maastricht framework. The Czech Republic and Slovakia have also started to impute the future fiscal cost of government credit guarantees into their deficit and debt figures reported under the Maastricht fiscal framework. Recent improvements in the coverage of general government deficit and debt reported by EU8 countries, however, fall short of fully capturing the future fiscal cost related to government risk-taking under PPPs. Specifically, take-or-pay contracts (including those signed by public utilities) and various types of guarantees that may be provided by the local as well as central levels of government under PPPs are not easily captured by ESA95, nor other accounting standards. Similarly, accountability structures in EU8 countries, although improving, fall short of ensuring fiscal prudence in the use and design of PPPs. EU8 countries have been improving the accountability of policy makers with respect to medium-term fiscal performance as they have been establishing medium-term fiscal frameworks and compliance with EU fiscal surveillance. EU8 countries have also been strengthening their audit mechanisms (internal audit by the MOF and external audit by the supreme audit institution) so as to promote accountability of policy makers for fiscal performance. The existing accountability frameworks in EU8 countries (as well as most other EU countries) are, however, still incomplete with respect to government risk taking and risk management. With respect to PPPs, policy makers do not seem accountable for the long-term fiscal risk arising from take-or-pay contracts and various types of guarantees offered by local and central governments. Similarly, there is no clear accountability for the adequacy of risk analysis that supports government decisions about fiscal support to infrastructure. Government accountability for managing risk exposures under PPPs is also limited. Information Good information on and understanding of the long-term fiscal cost of PPPs is important for promoting risk awareness (that is, an open discussion and acknowledgement of risks and government risk exposures). EU8 countries, however, have only limited information on the risks involved in PPPs and limited understanding of the long-term fiscal cost of PPPs. Moreover, these countries make very little of such information publicly available. PPP contracts and their content are considered confidential. This makes it difficult for policy analysts to assess the long-term fiscal cost of PPPs--and for the public to exercise appropriate pressure on policy makers for fiscal prudence. Without good information and understanding, the risks and long-term fiscal cost of PPPs tend to be underestimated. This further contributes to making PPPs and government risk-taking under PPPs an attractive choice of fiscal support to infrastructure. Capacity Building government capacity to evaluate and manage risks and long-term obligations that arise from PPPs takes time and effort. Weaknesses in government capacity to evaluate and manage risk may surface in the form of inefficient risk allocation and excessive government risk exposure under PPPs. Experience shows, however, that useful capacity can be built relatively quickly once the Ministry of Finance is committed to do so. EU8 countries have been building their fiscal risk management capacity. Most have centralized the authority for issuing government guarantees and for managing fiscal risk in their ministries of finance, have developed a central database of government guarantees and local government borrowing, and have built capacity to gather and analyze information on fiscal risk. Published analyses, such as Bezdek et al (2003), suggest that, for instance in the Czech Republic, the capacity to analyze fiscal risk (as well as disclosure) has been evolving. The capacity to manage government debt and its risks has also improved and, for instance in Hungary, is comparable to international good practice in terms of fine-tuning the instruments for the issuing of public debt, and the systematic use of benchmarking in order to minimize risk and costs (Currie, Dethier and Togo, 2003). The capacity to actively manage government risk exposures arising from contingent liabilities and control long-term obligations has been more limited. 6 4. Enhancing fiscal institutions for PPPs Fiscal institutions need to set the basis for transparency, accountability, and other measures to support fiscal prudence in the use and design of PPPs. It will be mainly a task for the domestic fiscal institutions to promote fiscal prudence in the use and design of PPPs. International mechanisms, however, may help if they better address government contingent liabilities and long-term obligations. In EU8 countries, EU fiscal surveillance in particular could become more effective in promoting fiscal prudence beyond the budget.13 Promote awareness of risk and long-term fiscal cost An open discussion of government long-term obligations and risk exposures enhances policy choices about fiscal support in infrastructure and improves government dealing with risk under PPPs. Similarly, at the level of local governments, introducing an open discussion and acknowledgement of risks, their sources, types, and possible long-term fiscal cost may deliver significant benefits in the soundness of local government involvement in PPPs. Most EU8 countries have been trying to collect, analyze, and discuss information about government risk exposures emerging from state credit guarantees. The Czech Republic, Estonia, and Latvia have been expanding the discussion to the whole portfolio of main government contingent liabilities and fiscal risk. This helps to build awareness about the risks and obligations under PPPs. Credible valuation of risks and assessment of long-term obligations contribute to risk awareness. Scenario analysis, for instance, has been useful to make policy makers aware of the potential fiscal impact of the worst possible outcomes under PPPs. Given the tendency of the proponents of PPPs to underestimate their long-term fiscal cost, impartial valuation based on credible assumptions and methodologies, conducted possibly by MOF, provides a useful basis for acknowledging risks and future possible fiscal cost. Impose disclosure Disclosure raises scrutiny, fiscal prudence, and the contestability of resources. Information that is disclosed invites scrutiny by people outside the government and by the government itself. When disclosure rules have broad coverage, they enable the government to better monitor lower-level governments and public sector units, and expand the share of government activities that is open to public scrutiny. Scrutiny is likely to generate pressure for greater fiscal prudence applied by governments at both the local and central levels. Modern financial reporting standards require the disclosure of commitments, contingent liabilities, and some other sources of fiscal risk. Thus, adopting such standards automatically creates a requirement to disclose information about most government risk exposures under PPPs. And, since government auditors must express an opinion on the accuracy of the disclosures, it automatically creates an enforcement mechanism. With respect to PPPs, governments can disclose their fiscal risks before as well as after PPP-deals are completed. By disclosing draft PPP contracts, the government would allow for a wider discussion of the fiscal risks involved and for a possible adjustment in the contract before it is signed. Disclosure of the signed PPP contracts would promote public understanding of government fiscal risk exposures, which in turn would promote government accountability with respect to PPPs. In addition, governments could be mandated to disclose their risk exposures under PPPs in the form of own description and analysis. Finally, to promote transparency and accountability, the fiscal cost of past PPPs should also be disclosed in a timely manner, possibly as part of a broader discussion on realized fiscal risk in annual budget documents. There are several prerequisites for both disclosure and risk awareness. These include having a database of government direct and contingent obligations to form a basis for analysis, adequate institutional 13This section draws on Brixi and Irwin (2004) and Brixi (2005). 7 capacity (including the capacity to gather and analyze relevant information and evaluate risk exposures), and an adequate enforcement mechanism (including a supportive political and legal environment, for instance, with respect to the reporting by local governments, public sector units, and public utilities) to ensure compliance. Given the long-term nature of fiscal costs of PPPs, it is appropriate to disclose long-term fiscal scenarios that would reflect the projected developments in government revenues and expenditures and would capture the possible fiscal effect of PPPs (along with other long-term obligations, such as those related to retirement and health care in aging societies). Enhance fiscal planning, accounting and budgeting For countries with large portfolios of contingent liabilities and long-term obligations, fiscal planning, to be meaningful, needs to reflect their possible long-term fiscal implications. Long-term fiscal projections discussed above would support government fiscal planning. Fiscal targets may be complemented by ceilings on government risk exposures, possibly in the simple form of ceilings on the face value of guarantees outstanding and of new guarantees issued in the coming budget year. Such simple ceilings have been in place, for instance, in Hungary, Latvia and Poland. In these countries, the ceilings are approved by the Parliament as part of the annual Budget Law. In Latvia, the annual Budget Law also provides ceilings for local government guarantees and borrowing. Furthermore, it specifies an allocation to cover the cost of past contingent liabilities that are considered likely to be realized in the given fiscal year. The ceilings, however, may not always be considered binding beyond a given year. In Poland, for instance, the guarantee ceilings fluctuated between 1.0% and 3.7% of GDP during 1999- 2003. Accounting reforms can address some but not all of the problems described here, in part because they can create requirements for routine, audited disclosure of some of the information described in the previous section. First, the adoption of "accrual" accounting, according to any reasonable standards, requires governments to report the assets they own, and a measure of the fiscal deficit that reveals approximate changes in the government's net worth, thus addressing the first reason why comparisons of PPPs and publicly financed investments are deceptive. Second, the adoption of some accrual accounting standards requires governments to report as liabilities some of the obligations created by PPPs. However, even the best existing accrual accounting standards are not sufficient for solving all the problems. Improvements in the analysis, disclosure, and management of the fiscal costs and risks of PPPs need not wait upon the adoption of better accounting and budgeting standards. Experience suggests that the benefits of greater scrutiny, cash neutrality, and risk awareness can be achieved with or without a comprehensive transition of the accounting and budgeting systems to accrual basis. Some countries (including the Czech Republic) have successfully combined reporting of contingent liabilities (and wider disclosure of risk) with cash accounting. Similarly, fiscal risk can be brought into the government's medium-term fiscal framework. This has already strengthened the accountability of policy makers and the quality of fiscal policy in a number of countries (including Hungary). Advance fiscal risk management to reduce government risk exposure The experience of governments trying to actively manage their risk exposures shows that fiscal risk management is very demanding. Governments find that to manage their risk exposures they need adequate information (hence a comprehensive database of all major risk exposures), the capacity to gather relevant information and the opportunity for open discussion, the ability to understand fiscal risk (which may be assisted by useful analytical frameworks), and incentives to act correctly--incentives that are supported by disclosure and adequate accounting and budgeting rules. Proper incentives in dealing with local government risk are supported by appropriate accountability structures. Policy makers need to be accountable for the adequacy of their risk analysis, assumptions, and decisions that involve fiscal risks and for managing the overall risk exposure of the government. Therefore, the role of the supreme audit institution (and the local audit bureaus) is to audit all aspects of government risk analysis and risk management. 8 Practice has shown the importance of three additional features of risk management: a clear risk management strategy (to specify to what extent the government is prepared to take on fiscal risk), centralized risk-taking authority (possibly in the budget office of the Ministry of Finance), and risk analysis and monitoring that are separate from risk taking (the debt management office and the supreme audit institution could be responsible for analyzing and monitoring risk internally and externally, respectively). The division of responsibilities and functions in risk management and the underlying reporting arrangements need to be very clear to provide a basis for adequate accountability structures. Reducing government risk exposure entails three complementary tasks: involving the private sector, transferring the risk to parties better able to bear it in the design of PPPs, and managing any residual risk that cannot be mitigated or transferred. Involving the private sector mainly implies mitigating the risk at the source and developing the financial markets. Ultimately, risk mitigation with private sector involvement is the most desirable long-run strategy. International mechanisms to promote fiscal prudence vis-à-vis PPPs In EU8 countries, risk awareness could be supported by expanding the framework of EU fiscal surveillance to involve a broad analysis of governments' fiscal positions, supported by surveys of risks arising in infrastructure, including those affecting local governments, public utilities, and other possible strategically important companies. The EU fiscal surveillance process could involve fiscal risk analysis and discussions with government officials that go beyond the government's own official statements. In the context of fiscal surveillance, it would be valuable to survey risks arising from infrastructure as well as risk exposures of local governments, exposures of state-controlled and strategically important companies (especially major suppliers of vital services) and various risk-prone financial institutions, such as credit and guarantee funds. Such surveys may promote the government's own understanding of its risk exposures and allow for a more accurate assessment of the country's fiscal performance. International mechanisms could do more to reward disclosure. Specifically, mechanisms, such as EU fiscal surveillance and IMF fiscal transparency assessment, need to reward countries that voluntarily expose the full scale of their fiscal risks, including risks under PPPs (that is, "upgrade" for transparency rather than "downgrade" for risks revealed). In assessing country fiscal performance, the positive value of transparency needs to be weighed carefully against the negative value of the revealed risks. Unfortunately, experience suggests that countries may be punished rather than rewarded when they reveal contingent liabilities. For instance, in 1997, when the Minister of Finance of the Czech Republic volunteered detailed information about until then unknown contingent liabilities and launched an effort to bring contingent liabilities under control, international institutions and sovereign credit rating agencies reacted with horror. To outweigh the cost of transparency, international mechanisms may need to do more to publicly praise countries for their steps toward transparency and disseminate analysis of actions taken by governments to control fiscal risks. 5. Conclusions EU8 countries would be well advised to raise the likelihood that PPPs will be used and designed well, and will deliver fiscal savings as well as promote investment in infrastructure. In this respect, EU8 countries could strengthen their fiscal institutions to set the basis for transparency, accountability and further measures to support fiscal prudence in the use and design of PPPs. Specifically, they would benefit from promoting awareness of risk and long-term fiscal cost, imposing disclosure, enhancing fiscal planning, accounting and budgeting, and advancing fiscal risk management to reduce government risk exposure. Institution building takes time, but international experience suggests that major advances in dealing with government risks and obligations of PPPs can be achieved when the government is committed to fiscal prudence. Government commitment to fiscal prudence (and to good fiscal performance overall), in turn, grows with better fiscal institutions. 9 International institutions could to a greater extent stimulate the commitment to change and support efforts to enhance fiscal institutions in EU8 countries. In particular, EU fiscal surveillance could be expanded to cover contingent liabilities, bring government contingent liabilities within the criteria by which country fiscal performance is assessed, and develop mechanisms to reward transparency and punish opacity and excessive risk taking. 10 References Bezdek, Vladimir, Kamil Dybczak and Ales Krejdl (2003), "Czech Fiscal Policy: Introductory Analysis." Working Paper series of the Czech National Bank. Czech National Bank. Prague (November). Black, Fischer, and Myron Scholes (1973), "The Pricing of Options and Corporate Liabilities." Journal of Political Economy. Volume 81. Pages 637­654. Brealey, Richard A., I. A. Cooper, and M. A. Habib (1997), "Investment Appraisal in the Public Sector." Oxford Review of Economic Policy. Volume 13. Issue 4. Pages 12­28. Brixi, Hana Polackova (2005), "Contingent Liabilities in New Member States." In Fiscal Surveillance in EMU: New Issues and Challenges, ed. European Commission. Edward Elgar Publishing (forthcoming). Brixi, Hana Polackova and Timothy Irwin (2004), "Fiscal Support for Infrastructure: Toward a More Effective and Transparent Approach." World Bank (processed). Brixi, Hana Polackova and Ashoka Mody (2002), "Dealing with Government Fiscal Risk: An Overview." In Government at Risk: Contingent Liabilities and Fiscal Risk ed. by Brixi and Schick. Washington DC: World Bank and Oxford University Press. Brixi, Hana Polackova, and Allen Schick (editors) (2002), "Government at risk: Contingent liabilities and fiscal risk." World Bank and Oxford University Press. Buiter, Willem (1990), "Principles of Budgetary and Financial Policy." Burns and Yoo (2002), "Public Expenditure Management in Poland" OECD, Economics Department Working Papers No. 346. CASE (2005), "Background Paper on Fiscal Risks from PPPs" Mimeo, Warsaw, Poland. Currie, Elizabeth, Jean-Jacques Dethier, and Eriko Togo (2003), "Institutional Arrangements for Public Debt Management." Policy Research Working Paper WPS3021. World Bank. EBRD Transition Report (2004). Ehrhardt, David, and Timothy Irwin (2004), "Avoiding Customer and Taxpayer Bailouts in Private Infrastructure Projects: Policy Toward Leverage, Risk Allocation, and Bankruptcy." World Bank Policy Research Working Paper 3274. April. European Commission (2001), "European Transport Policy for 2010: Time to Decide." European Commission (2004), "Resource book on PPP case studies." Directorate-General, Regional Policy. European Commission (2005), "Assessments of the December 2004 Updates of Convergence Programs of EU8 countries." Eurostat (2004), News release No. 18, February 11, "Treatment of Public Private Partnerships." Available via the Internet: http://europa.eu.int/comm/eurostat/Public/datashop/print- product/EN?catalogue=Eurostat&product=2-11022004-EN-AP-EN&mode=download Gómez-Ibáñez, José A. 1997. Mexico's private toll road program. Kennedy School of Government Case Program, Harvard University. IMF (2004a), "Public-Private Partnerships" Washington DC. IMF (2004b), "Government Finance Statistics Yearbook". Irwin, Timothy (2003), "Accounting for Public-Private Partnerships: How Should Governments Report Guarantees and Long-Term Purchase Contracts?" Irwin, Timothy (2004), "Measuring and Valuing the Risks Created by Revenue and Exchange-Rate Guarantee in Korea." Chapter 3 of Developing best practice for Korea's PPI market: With a focus on PSC. Private Infrastructure Investment Center of Korea. 11 Irwin, Timothy, Michael Klein, Guillermo Perry, and Mateen Thobani (editors) (1997), "Dealing with Public Risk in Private Infrastructure." World Bank. PriceWaterhouseCoopers (2004), "Developing Public Private Partnerships in New Europe." Quiggin, John (2004), "Risk, PPPs, and the Public Sector Comparator." Australian Accounting Review. July. Volume 14. Number 2. Pages 51­61. Westwood, J. N. (1964), "A History of Russian Railways." London: George Allen and Unwin Ltd. World Bank (2004), "World Development Report: A Better Investment Climate for Everyone." New York: Oxford University Press. 12 ANNEX. Table 1. Selected liabilities from PPPs and other privately financed infrastructure projects14 Direct Contingent Obligation in any Obligation if a particular event occurs event Explicit Obligations to State guarantees of revenues under PPPs make Poland (A1 motorway) Government availability liability payments for Hungary (M5 motorway) created by a services State guarantees of debt of PPPs law or Hungary Poland ­ A2 motorway, energy project in Krakow, water and sewage treatment in Warsaw and Torun, road and rail infrastructure in Lodz. contract (prison, educational Statutory guarantees on liabilities and other obligations of various entities, facilities including financial institutions, possibly involved in PPPs (infrastructure and the M6 development funds, etc) motorway) Czech Republic (Railway Transport Infrastructure Administration) Tax Hungary (State Development Bank) expenditures,li Poland (Bank for Environment Protection providing guaranteed credit lines for ke exemptions small and medium enterprises and programs co-financed by EU) Poland (tax Estonia, Latvia, Lithuania, and Slovakia (investor protection guarantees) exemptions State guarantees on service purchase contracts for state- owned Poland and Hungary (possible obligations arising from the past power-purchase companies) agreements) Litigation Slovakia (legal claims by CSOB and the Slovak Gas Company) Implicit Not typical in Possible assumption of the obligations of state-owned utilities that have entered PPPs; would into unguaranteed power-purchase agreements with independent power A "political" arise if producers obligation of governments Possible assumption of the debts of privately financed toll roads, beyond the government were politically, extent of formal guarantees that reflects but not legally, Hungary (M1/M15 motorway construction concessions ­ partly implemented public and through the Road Construction Corporation of the State Development Bank) obliged to interest-group Poland (claims arising from expressway construction concessions) continue to pressures provide subsides Possible claims by local governments to assist in covering their own debt, guarantees, arrears, letters of comfort and other obligations possibly related to to a PPP. PPPs Poland (local government debt and guarantees related to regional development) Czech Republic (bail-outs related to hospital arrears) Claims by financial institutions involved in PPPs Slovenia (regional guarantee schemes) Note: The table shows the obligations of the general central government (rather than the consolidated public sector). Source: The authors, using the framework set out in Polackova (1998). 14Table 1 contains only selected examples of fiscal liabilities and risks from PPPs in EU8 and is by no means comprehensive. 13 Table 2. Improving fiscal institutions for PPPs at a glance Goal Options Time horizon Risk Collect and centralize information on PPP contracts S awareness Discuss risks and long-term fiscal cost of PPPs as part of government decision- S making M Promote valuation of risks and obligations Disclosure Disclose past fiscal cost of PPPs S Disclose outstanding PPP contracts S Disclose government analysis of risks and long-term obligations under PPP S-M contracts S-M Disclose draft PPP contracts and government analysis of related risks and M obligations M Disclose long-term fiscal scenarios reflecting the future possible fiscal effect of PPPs Enhance the whole system of financial reporting standards to require disclosure of commitments, contingent liabilities and other sources of fiscal risk in the public sector Accounting, Reflect the net present value of expected fiscal cost of PPPs in government S budgeting, deficit and debt when government obligation originates and fiscal Reflect the future possible fiscal effect of PPPs in fiscal planning S planning Set overall limits on government risk exposure--either as simple ceilings on the S face value of government guarantees or as part of joint ceilings on cash outlays and the discounted cost of guarantees Consider reducing deficit/debt ceilings by risk-adjusted values of contingent S-M liabilities issued/outstanding and/or establishing contingent liability fund Enhance the whole system accounting and budgeting standards to address M government commitments, contingent liabilities and other sources of fiscal risk on a systematic basis Risk Monitor government risk exposures and obligations S managemen Centralize government risk-taking authority (except where the accountability of S t decentralized organizations is strong enough) Audit government risk analysis and risk management S Build capacity to evaluate and manage risk S-M Develop extended assets and liabilities management framework M Note: S = short term, M = medium term. 14