Three Policy Options for Crown Financial Policy

Hansen, Eric Working Paper Three Policy Options for Crown Financial Policy New Zealand Treasury Working Paper, No. 03/30 Provided in Cooperation with: The Treasury, New Zealand Government Suggested Citation: Hansen, Eric (2003) : Three Policy Options for Crown Financial Policy, New Zealand Treasury Working Paper, No. 03/30, New Zealand Government, The Treasury, Wellington This Version is available at: https://hdl.handle.net/10419/205535 Standard-Nutzungsbedingungen: Die Dokumente auf EconStor dürfen zu eigenen wissenschaftlichen Zwecken und zum Privatgebrauch gespeichert und kopiert werden. Sie dürfen die Dokumente nicht für öffentliche oder kommerzielle Zwecke vervielfältigen, öffentlich ausstellen, öffentlich zugänglich machen, vertreiben oder anderweitig nutzen. Sofern die Verfasser die Dokumente unter Open-Content-Lizenzen (insbesondere CC-Lizenzen) zur Verfügung gestellt haben sollten, gelten abweichend von diesen Nutzungsbedingungen die in der dort genannten Lizenz gewährten Nutzungsrechte. Terms of use: Documents in EconStor may be saved and copied for your personal and scholarly purposes. You are not to copy documents for public or commercial purposes, to exhibit the documents publicly, to make them publicly available on the internet, or to distribute or otherwise use the documents in public. If the documents have been made available under an Open Content Licence (especially Creative Commons Licences), you may exercise further usage rights as specified in the indicated licence. https://creativecommons.org/licenses/by/4.0/ Three Policy Options for Crown Financial Policy Eric Hansen N EW Z EALAND T REASURY W ORKING P APER 03/30 D ECEMBER 2003 Treasury:553144 NZ TREASURY WORKING PAPER 03/30 Three Policy Options for Crown Financial Policy MONTH / YEAR December 2003 AUTHOR Eric Hansen New Zealand Treasury PO Box 3724 Wellington NEW ZEALAND Email Telephone Fax [email protected] 64 4 472 2733 64 4 473 0982 ACKNOWLEDGEMENTS I wish to thank John Carran, Philip Combes, Aaron Gill, Arthur Grimes, Greg Horman, Brian McCulloch and Tim Ng for comments on an earlier draft of this paper. All errors remain my responsibility. NZ TREASURY New Zealand Treasury PO Box 3724 Wellington 6008 NEW ZEALAND Email Telephone Website [email protected] 64-4-472 2733 www.treasury.govt.nz DISCLAIMER This paper was written while the author was on the staff of the New Zealand Treasury. The views expressed in this Working Paper are those of the author(s) and do not necessarily reflect the views of the New Zealand Treasury. The paper is presented not as policy, but with a view to inform and stimulate wider debate. WP 03/30 | THREE POLICY OPTIONS FOR CROWN FINANCIAL POLICY i Abstract Crown financial policy is concerned with how the government manages the Crown’s assets and liabilities. The recently established New Zealand Superannuation Fund, which is projected to grow to around 45% of GDP over the next few decades, highlights that Crown financial policy is likely to become an important economic policy tool with potentially significant implications for New Zealand economic welfare. Previous work has identified that four objectives should form the main basis for assessing alternative Crown financial policies. Three of the objectives relate to minimising distortionary taxation, time-inconsistency of policy and agency cost of government. However, the absolute and relative importance of these objectives is subject to considerable uncertainty. The fourth objective, which is to avoid exacerbating any existing inefficiencies or creating any new ones, is considered part of the baseline common to all policies. In this paper a qualitative analysis is conducted to select three high-level policies for detailed quantitative analysis in future papers. The three policies vary in terms of level of risk: • A low risk policy that places emphasis on time-consistency and agency cost issues while down-weighting the significance of distortionary taxation; • A medium risk policy that applies a balanced weighting to the three issues; and • A high risk policy that places emphasis on distortionary taxation while downweighting time-consistency and agency cost. Detailed policy targets are specified for the candidate policies in terms of Crown net worth, overall risk/return properties of the Crown balance sheet, and the level and structure of financial assets and public debt. The policy targets under the status quo are presented for comparative purposes. JEL CLASSIFICATION E61 H11 H63 KEYWORDS Agency cost of government; Bayesian decision theory; comparative institutional method; Crown balance sheet; public debt management; distortionary taxation; time-inconsistency of policy WP 03/30 | THREE POLICY OPTIONS FOR CROWN FINANCIAL POLICY ii Table of Contents Abstract ...............................................................................................................................i Table of Contents ..............................................................................................................ii List of Tables......................................................................................................................ii List of Figures...................................................................................................................iii 1 Introduction ..............................................................................................................1 2 Objectives and targets.............................................................................................3 2.1 Main objectives ...............................................................................................................3 2.2 Conflicting targets ...........................................................................................................7 3 Framework for analysis ...........................................................................................8 3.1 Comparative institutional method ...................................................................................8 3.2 Uncertainty and policy making........................................................................................9 3.3 Qualitative risk assessment..........................................................................................13 4 Assessment of policy options...............................................................................14 4.1 Multiple objectives ........................................................................................................15 4.2 Key judgements ............................................................................................................16 4.3 Low risk policies............................................................................................................17 4.4 Medium risk policies .....................................................................................................18 4.5 High risk policies...........................................................................................................19 4.6 Summary of proposed policy options ...........................................................................20 5 Detailed policy targets ...........................................................................................21 6 Conclusions............................................................................................................24 References .......................................................................................................................26 Appendix I: Policy making under uncertainty..............................................................29 Appendix II: Qualitative Assessment of Risks..............................................................32 Part A: General Discussion.....................................................................................................32 Part B: Application to Loss Functions ....................................................................................38 List of Tables Table 1 - Policy targets under status quo and four objectives ............................................................6 Table 2 - Available policy options .................................................................................................15 Table 3 - Low risk policies.................................................................................................................17 Table 4 - Medium risk policies .......................................................................................................18 Table 5 - High risk policies .............................................................................................................20 Table 6 – Status quo and candidate policy options ........................................................................23 Appendix Table 1 – Available policy options with three objectives..................................................38 Appendix Table 2 – Matrix of loss functions with three objectives...................................................38 Appendix Table 3 – Number of “High (H)” assessments .................................................................51 WP 03/30 | THREE POLICY OPTIONS FOR CROWN FINANCIAL POLICY iii List of Figures Figure 1: Conflicting targets ................................................................................................................7 Figure 2: Decision structure facing policy maker ..............................................................................12 Figure 3: Efficient policy frontier........................................................................................................13 Figure 4: Policy options and losses...................................................................................................16 WP 03/30 | THREE POLICY OPTIONS FOR CROWN FINANCIAL POLICY 1 Three Policy Options for Crown Financial Policy 1 Introduction Crown financial policy specifies how the government manages the Crown’s assets and liabilities.1 Policy analysis in this area is concerned with how the structure and size of the Crown balance sheet could affect the decisions of citizens in managing their own wealth portfolios and government decisions on fiscal and other economic policies. The practical effect of movements in the Crown balance sheet was evident in the Crown Financial Statements for 2002/03. Due to accrual accounting, a partial revaluation of the Crown balance sheet in 2002/03 reduced the government’s operating surplus from $4 billion to $1.4 billion. This adjustment amounted to around 6% of Core Crown revenue.2 Crown financial policy is likely to become progressively more important over time as the recently established New Zealand Superannuation Fund accumulates financial assets equivalent to 45% of GDP or around $56 billion in current terms.3 If these funds accumulate as projected, then a 10 basis point (or 0.1%) improvement (decline) in annual returns at the same risk level would confer a net present value gain (loss) to New Zealand of around $1 billion (at 5% discount rate). Hansen (2003) organises the theoretical literature within a policy framework to identify specific objectives, targets and instruments associated with the main theories.4 The paper views the Crown balance sheet as a policy instrument and identifies seven potential objectives to which Crown financial policy could be targeted. 1 This paper uses the term Crown financial policy to mean government policies relating to the management of the Crown’s aggregate balance sheet. The Crown balance sheet includes the Crown’s ownership interest in state-owned enterprises and other central government assets and liabilities meeting Generally Accepted Accounting Practice (GAAP) but excludes Local Authority assets and liabilities. A wider definition of Crown financial policy would include measurement issues, financial reporting and performance and accountability issues but these are excluded for the purposes of this paper. 2 See Crown Financial Statements at http://www.treasury.govt.nz/. 3 McCulloch and Frances (2001) describes the New Zealand Superannuation Fund. 4 The Treasury has conducted research on Crown financial policy in one form or another since at least the mid-1990s. Skilling (1997) and Davis (2001) summarise and develop the literature relevant to Crown financial policy, while Grimes (2001) discusses the operational objectives and practices relevant to managing the Crown’s balance sheet. Empirical analyses by Huther (1998), Fabling (2002) and Davis and Fabling (2002) have found tentative evidence that it may be possible to improve the performance of the Crown balance sheet. International contributions include Bohn (1990, 1995), Chari, Christiano and Kehoe (1994), Leong (1999), Lucas and Stokey (1983), and Missale (1997, 1999). WP 03/30 | THREE POLICY OPTIONS FOR CROWN FINANCIAL POLICY 2 The objectives considered most important related to distortionary taxation, timeinconsistency of policy and agency cost of government. A fourth objective, called “downside efficiency risk”, relating to the need to avoid exacerbating existing inefficiencies or creating new ones, was also identified as relevant to the setting of policy targets. Section 2 provides a summary of these objectives and associated policy targets. Purpose The purpose of this paper is to fashion the above objectives into several alternative policies suitable for future quantitative analysis. In essence, the paper looks ahead to consider how policy makers would be likely to choose between alternative policy options and the requirements they would have. The aim is to provide direction to ensure Treasury’s future research is focussed on producing evidence relevant to policy decisions and avoid spending resources on comparing policies that actually would not be implemented. The approach adopted in this paper draws from Coase’s comparative institutional method and Bayesian decision theory. The comparative institutional method emphasises that policy analysts should avoid comparing real world policy proposals (which necessarily are imperfect) against idealised textbook models. Rather, any policy proposal should be compared against the counterfactual of what would be most likely to happen in the absence of the proposed policy. A critical fact is that the choice between policy options will be made in a world of uncertainty. Irrespective of how much research is undertaken, policy makers will be faced with considerable uncertainty about the true model of the economy and therefore the relative importance of distortionary taxation, time-inconsistency of policy and agency cost of government. Recognising these uncertainties, a detailed qualitative analysis of alternative policies is conducted based on Bayesian decision theory. Structure The structure of the paper is as follows. The next section (Section 2) provides brief descriptions of the main policy objectives identified in Hansen (2003) and the detailed policy targets implied by these objectives. Section 3 discusses the framework adopted in this paper based on comparative institutional and Bayesian decision theory. Section 4 applies the framework to select three high-level policies representing low, medium, and high risk. Section 5 specifies the three candidate policies in terms of the policy targets that would apply in each case. Conclusions are presented in Section 6. WP 03/30 | THREE POLICY OPTIONS FOR CROWN FINANCIAL POLICY 3 2 Objectives and targets This section summarises the study of objectives, targets and instruments in Hansen (2003) and discusses potential conflicts between targets. The status quo policy is also discussed for comparative purposes. 2.1 Main objectives The study in Hansen (2003) concluded that four objectives should inform the design of policy options for Crown financial policy. This section concludes that for the purposes of this paper the focus should be on three objectives – relating to distortionary taxation, timeconsistency and agency cost – and that downside efficiency risk (the fourth objective) may be considered part of the baseline common to all candidate policies. Distortionary taxation Taxes, due to their involuntary nature, create incentives for taxpayers to substitute away from taxed activities toward activities that are not taxed, or are taxed at lower marginal rates. If the taxed activities would otherwise be worthwhile, the substitution reduces welfare and creates a deadweight loss. A possible policy objective is to minimise the deadweight losses of taxation subject to satisfying the Crown’s inter-temporal budget constraint (IBC).5 A clear conclusion from all tax smoothing models studied is that all diversifiable risk should be eliminated. Minimising deadweight losses also implies smoothing of the tax rate over the economic cycle and pre-funding of an anticipated permanent increase in government spending (called deterministic smoothing). Other possible policy targets are conditional on particular assumptions: • If citizens are fully rational and not constrained by imperfect capital markets, policy should target the minimum-variance portfolio, i.e. in addition to eliminating diversifiable risk, policy should minimise systematic risk up to the maximum extent permitted by available instruments.6 5 The inter-temporal budget constraint requires that at any date the sum of net worth as at that date plus net present value of future tax revenue be greater than or equal to the net present value of future government spending. 6 If citizens do not alter their portfolios optimally in response to changes in the Crown portfolio, the appropriate policy may involve targeting a level of systematic risk greater than the minimum feasible level. The systematic risk of a portfolio is the risk that cannot be avoided by diversifying the portfolio across the risky assets available in the (global) economy, so that returns on the portfolio will vary with the economy (Copeland and Weston 1988). The systematic risk of a portfolio can be altered by increasing or decreasing the proportion of the portfolio invested in the safe asset (proxied by government bonds). WP 03/30 | THREE POLICY OPTIONS FOR CROWN FINANCIAL POLICY 10 The key problem with the Neyman-Pearson method is that it assumes the cost of falsely rejecting the status quo (the null hypothesis) is an order of magnitude larger than the cost of falsely rejecting the competing proposal (the alternative hypothesis).19 While this assumption may be appropriate for guiding the progress of science, for policy (and other decisions) the relative magnitude of costs often are more balanced. To provide a sound basis for policy to maximise economic welfare, it is necessary to estimate of the costs and benefits arising under various scenarios and the probabilities of those scenarios, occurring. Blaug (1992 pp.21-3) and Gorringe (1991, 1998) discuss these issues in more detail. An example where the inappropriateness of statistical inference for policy making has been taken seriously is the literature on monetary policy under model uncertainty. The monetary policy literature assesses the performance of alternative policy reaction functions in circumstances where the “true” model of the economy is different from that specified in the reaction function.20 Bayesian decision theory In addition to comparative institutional analysis, we adopt a version of Bayesian decision theory. In essence, the Bayesian approach is an application of standard micro-theoretic decision making under uncertainty where the decision maker’s objective is to maximise expected utility.21 A key assumption in the Bayesian approach is that the decision maker can assign a subjective probability to every potential outcome.22 To develop notation for later use, consider a policy option (PO) that is an alternative to the status quo (or another policy option). The policy maker must decide whether to implement the policy option or retain the status quo in the face of uncertainty about the true nature of the economy. A decision to implement is indicated by placing a high weight (denoted “H”) on the policy option, whereas a decision to retain the status quo is indicated by placing a low weight (“L”) on the policy option.23 19 A second problem is that in applications the null hypothesis usually specifies the variable of interest as having “zero coefficient”. The statistical test conducted is whether the coefficient is zero (null hypothesis) or non-zero (alternative hypothesis). However, in most policy decisions the relevant test is whether the coefficient is less than or equal to zero versus positive. 20 See, for example, Christodoulakis, Kemball-Cook and Levine (1993) and Onatski and Williams (2003) for general discussions and Conway et. al. (1998) and Drew and Hunt (1999) for analyses relevant to monetary policy in New Zealand. 21 Descriptions of Bayesian decision theory and applications are available in Cyert and DeGroot (1987), Gorringe (1991, 1998), Hirshleifer and Riley (1992), Rhodes (1994), and Silvey (1975). 22 Knight (1920) distinguished between risk and uncertainty on the basis of the existence of numerical probabilities. He defined a state of uncertainty as existing for an event when no numerical probability or frequency of the event occurring can be assigned. In contrast, a state of risk exists when a numerical probability can be assigned. Bayesian decision theory is based on the view that every event can be assigned a subjective probability by the decision maker, so that no distinction is made between risk and uncertainty. 23 In Section 5 the high and low weights correspond to the assignment of different values to policy targets. A high weight corresponds to assigning values to the relevant policy targets at the conservative end of plausible values, implying strong constraints on the Crown balance sheet, while a low weight corresponds to assigning values that imply weaker constraints. The approach is flexible in that the values assigned could represent the long-term direction of policy (as assumed in Section 5) or could be small changes from the status quo for the purpose of conducting marginal analysis of the next step along a transition to the long-term policy target. Also, the assumption of two states and two policy choices is a simplification. The references in Footnote 21 provide more general analyses where the state space and decision choices are continuous. WP 03/30 | THREE POLICY OPTIONS FOR CROWN FINANCIAL POLICY 11 Suppose that the “correct” decision depends on whether the true impact of the policy action exceeds a threshold value. If the impact exceeds the threshold it is called economically significant (“s”) and the policy option should be implemented. If the impact is less than the threshold it is called economically insignificant (“i”) and the status quo should be retained. The policy maker does not know whether the impact would be significant or insignificant but has formed a probability distribution over the true states, s and i. There are four possible outcomes: • True Positive (TP) The impact is significant (s) and the policy maker correctly assigns high weight (H); or • False Negative (FN) The impact is significant (s) but the policy maker incorrectly assigns low weight (L); or • True Negative (TN) The impact is insignificant (i) and the policy maker correctly assigns low weight (L); or • False Positive (FP) The impact is insignificant (i) but the policy maker incorrectly assigns high weight (H).24 Let the payoffs for the scenarios be random variables VTP, VFN, VTN and VFP. The structure of the decision and possible outcomes are illustrated in Figure 2 below, where the dashed oblong indicates the policy maker (P) does not know whether the true state is ‘s’ or ‘i’. The true state is thought of as being determined by nature (N). A property of a decision structure of this nature is that attention can be restricted to the false positive and false negative decision errors. This is achieved by defining the loss from a decision error as the value forgone relative to the correct decision.25 Bayesian decision theory assumes there exists a utility function that represents the decision makers’ risk preferences over losses incurred under false positive and false negative errors. Consistent with standard micro-economic theory, the optimal decision is to implement the policy option if and only if the expected utility of losses under implementation is less than the expected utility of losses under the status quo. 24 In statistical theory the false positive is called a Type I error and the false negative is called a Type II error. The former error is the decision to reject a null hypothesis that is in fact true while the later error is the decision to accept a null hypothesis that is in fact false. 25 The losses are defined as LFN = VTP – VFN and LFP = VTN – VFP. WP 03/30 | THREE POLICY OPTIONS FOR CROWN FINANCIAL POLICY 12 Figure 2: Decision structure facing policy maker Application to public policy The practice of applying Bayesian decision theory to public policy needs to take into account that the policy analyst conducting the analysis is not the same person as the policy maker. In particular, the policy analyst is unlikely to know the policy maker’s prior probability beliefs and utility function (i.e. degree of risk aversion). One approach for overcoming the information gaps could involve surveying the policy maker’s prior beliefs and risk preferences and then proceeding to derive the optimal decision. Another approach, adopted in this paper, would involve the policy analyst presenting to the policy maker a menu of policy options that differ in their risk levels. The policy maker would compare expected loss and measures of risk across policy options, and would choose the policy with risk/return properties that best meets his or her preferences. The policy options may be ordered from low to high risk, drawing out an efficient frontier of policy options, as illustrated in Figure 3 below. Policy options lying inside the frontier would be dominated and could be discarded. If the policy maker’s preferences are transitive, the most desired policy option may be identified through a sequence of pair-wise tests where the policy maker assesses whether the incremental benefit offered by policy option ‘n’ relative to policy option n-1 is outweighed by the increased risk: Policy Option 2 tested against Policy Option 1, i.e. Test: PO2 ≤ PO1 Policy Option 3 tested against Policy Option 2, i.e. Test: PO3 ≤ PO2 … Policy Option N tested against Policy Option N-1, i.e. Test: PON ≤ PON-1, where “≤ ” is read as meaning “less preferred than”. s i N L VTN VFP H P 0LFP H VTP VFN L P Outcomes: 0 LFN Losses: WP 03/30 | THREE POLICY OPTIONS FOR CROWN FINANCIAL POLICY 13 Figure 3: Efficient policy frontier 3.3 Qualitative risk assessment A full evaluation of the risk properties associated with each policy would require the conduct of comprehensive empirical research. The purpose of this section is to identify qualitative indicators to distinguish risk levels in broad terms in advance of more detailed empirical analysis. Average, variance and worst-case losses A risk averse policy maker would be concerned about the probability of a decision error, the average loss and range of losses possible if such an error did occur. Relevant considerations would include the degree of uncertainty about the level of losses and the magnitude of ‘worst-case’ losses. Persistence vs. reversibility The policy decision may or may not be subject to future review. A policy where a decision error would be revealed early in the post-implementation period and which could be reversed quickly and at low cost normally would be less risky than a policy that would be irreversible or reversible only slowly and at high cost. An example of the difference is provided by the different roles of the Commerce Commission in Mergers & Acquisition applications versus some other regulatory functions. M&A decisions are clearly irreversible while decisions whether or not to impose price regulation are reversible. In addition, a decision to impose price regulation may stifle the release of further information about the competitiveness of the market whereas a decision to not impose price regulation has the advantage of allowing further observations about competitive conditions and thus whether regulation is warranted. Monetary policy, where decisions on the Official Cash Rate are made six-weekly, is another example where policy decisions are made under considerable uncertainty but have the benefit of being reversible. PO 1 Efficient frontier Risk Inverse of expected losses PO 2 PO 3 Dominated PO WP 03/30 | THREE POLICY OPTIONS FOR CROWN FINANCIAL POLICY 14 Combinations of objectives In the case of multiple objectives, the impact of placing high or low weight on each objective depends on which other objectives are included in the policy. For example, in the case of Crown financial policy, a policy that places high weight on both tax-smoothing and agency cost would have quite different implications than if either objective received high weight alone. The former policy is balanced in the sense of creating countervailing incentives for and against the build up of financial assets, while the other policies that place high weight on one objective would result in a more extreme balance sheet structure. Thus, in most cases, the risks associated with an objective can be evaluated only by taking into consideration the other components of policy. Implementation risk Proposed policy options must be capable of being implemented without undue risk. One source of implementation risk arises when the policy is based on a model that assumes decision makers have more information and better capability to interpret that information than would actually be the case. This concern with implementation risk is consistent with Coase’s dictum noted above that economists should always judge alternative arrangements as they would actually operate. Voter misperception risk Voter misperception risk is defined as the risk that an economically desirable decision leads to outcomes that the public perceives as a mistake. For example, in the case of Crown financial policy, a desirable hedging strategy could lead to outcomes where the market value of Crown financial assets were revised downward by (say) $5 billion as an offset to upward revaluation in the tax asset. Because the Operating Balance would include the downward revision in the financial asset but exclude the increase in the tax asset, the public may view the outcome as reflecting poor economic management by the government of the day. Misperception risks drive a wedge between the interests of policy makers (as agents) and the public (as principals), potentially causing desirable policy to be over-turned in favour of a false positive or false negative position. 4 Assessment of policy options This section applies the framework developed above to conduct a qualitative assessment of stylised policy options available for Crown financial policy. Eight policies are identified and assigned either to the low, medium or high risk category (Appendix II provides the detailed qualitative analysis in support of these assignments). From each risk category, one policy option is chosen as warranting detailed empirical analysis. The section begins with an outline showing how the framework is applied to Crown financial policy issues. This is followed by a discussion of key judgements made and then an analysis of each risk category. WP 03/30 | THREE POLICY OPTIONS FOR CROWN FINANCIAL POLICY 15 4.1 Multiple objectives The framework developed in Section 3.2 considered the case where the policy choice is defined by a single variable. The analysis assumed that the policy maker chooses whether to place high (“H”) or low (“L”) weight on the proposed policy in face of uncertainty about whether the impact on economic welfare would be significant (“s”) or insignificant (“i”). The application to Crown financial policy involves extending the framework to a multi-variate analysis. This reflects that the Crown balance sheet may impact on economic welfare through three main channels: • Deadweight losses caused by the average level and variability of the tax rate; • Higher debt servicing costs and reduced investment due to greater uncertainty (and losses caused by confidence crises) that could result if monetary or fiscal policy became time-inconsistent; and • Inefficient forms of government expenditure arising as a result of the principal-agent relationship between government and citizens. Both the absolute and relative magnitudes of these effects are subject to considerable uncertainty. In the following the true state or “true model” of the economy is described by the vector (x, x, x), where each “x” may be significant (s) or insignificant (i). For example, the state vector (s, i, i) would indicate that tax smoothing is economically significant but losses due to time-inconsistency and agency cost are insignificant. Similarly, policy is described by the vector (X, X, X) where each “X” may be high (H) or low (L) weight. For example, the policy option (H, L, L) would indicate high weight on policy targets aimed at tax smoothing and low weights on policy targets to mitigate timeinconsistency and agency cost. This structure gives rise to eight (= 23) possible policy options, as listed in Table 2. Table 2 - Available policy options Policy Options Distortionary taxation Timeconsistency Agency cost 1 2 3 4 5 6 7 8 H H H L H L L L H H L H L H L L H L H H L L H L WP 03/30 | THREE POLICY OPTIONS FOR CROWN FINANCIAL POLICY 16 Figure 4 illustrates that for each policy there are eight possible outcomes that depend on the true model of the economy. In the case where the policy vector matches the state vector the loss is zero. In all other cases the losses are positive. Figure 4: Policy options and losses 4.2 Key judgements A key judgement made in this paper is that existing institutional arrangements do not fully mitigate the risks of time-inconsistency and agency cost in all relevant circumstances. Although the risk of time-inconsistency appears low in the context of the Reserve Bank Act 1989 and Fiscal Responsibility Act 1994 and currently moderate debt levels, the level of risk could increase significantly if the Crown followed a highly leveraged strategy in pursuit of tax smoothing. Empirical evidence on tax smoothing for both NZ and the US imply debt levels in excess of 2500% of GDP (Bohn 1990 and Davis and Fabling 2002). It may be unwise to assume the current legislative arrangements would be robust to a highly leveraged strategy.26 In terms of agency cost, current institutional arrangements such as the New Zealand Superannuation Act 2001 are likely to help protect against “direct raiding” of funds by a future government. However, these arrangements do not preclude “indirect raiding” or “expenditure creep”, whereby the pressure for inefficient government expenditure would increase as Crown net worth increases. Related to these issues, it is assumed that a high debt target would be an effective instrument in reducing agency cost. The proposition has been adopted as an application of Jensen’s (1986) free cash flow theory of corporate finance. 26 Future analysis could assess the possibility of strengthening existing arrangements or adding new arrangements. For example, adopting a foreign currency in place of the New Zealand dollar would remove governments ability to engage in surprise inflation. (H, H, H) PO1 (H, H, L) PO2 (H, L, H) PO3 (L, H, H) PO4 (H, L, L) PO5 (L, H, L) PO6 (L, L, H) PO7 (L, L, L) PO8 (s,s,s) LFN,TP,TP (s,s,i) LFN,TP,FP (s,i,s) LFN,FP,TP (i,s,s) LTN,TP,TP = 0 (s,i,i) LFN,FP,FP (i,s,i) LTN,TP,FP (i,i,s) LTN,FP,TP (i,i,i) LTN,FP,FP WP 03/30 | THREE POLICY OPTIONS FOR CROWN FINANCIAL POLICY 17 However, the inter-temporal budget constraint facing the Crown is usually thought of as less stringent than for a corporation. Except for the recent period in New Zealand, it is not clear that high levels of debt has been an effective constraint on government spending in many countries.27 A further key judgement is that the potential loss from a false positive error on distortionary taxation is significantly larger than for a false positive error on timeinconsistency and agency cost. The difference is that placing high weights on the timeconsistency and agency cost issues serves to constrain government policy action while a high weight on distortionary taxation would motivate an “activist” policy. Actively seeking to smooth taxes carries the significant risk of implementation failure due to highly uncertain correlations between asset returns. A further risk, due to the taxation of capital, is that tax smoothing could lead citizens to take on excessive risk in their portfolios (Coleman, 1997). Thus, incorrectly placing high weight on tax smoothing (a false positive error) would carry the risk of large losses in the absence of other constraints. The judgement is made that a high level of debt would attract a risk premium for default risk. It is assumed that the risk premium would be “unjustified” in the sense that the country never intends to default or deflate the real value of the debt. The result of this judgement is that both distortionary taxation and time-consistency objectives motivate an upper bound on debt levels. Therefore, the two objectives tend to be substitutes for each other when evaluating the losses associated with debt targets. A further working assumption made throughout this section (and Appendix II) is that the debt target implied by the agency cost objective conflicts with the debt target implied by the time-consistency and distortionary taxation objectives. This need not be the case as the “high” debt target motivated by agency cost considerations could be lower than the “low” debt target motivated by the other two objectives. 4.3 Low risk policies The analysis presented in Appendix II suggests that low risk policies are those that place low weight on tax smoothing and avoid conflict between agency cost and timeconsistency. Table 3 shows two such policies and a third where agency cost and timeconsistency are conflicting.28 Table 3 - Low risk policies Policy Options Distortionary taxation Timeconsistency Agency cost 4 6 8 L L L H H L H L L 27 This issue deserves further analysis. If high debt was judged to be an ineffective instrument for reducing agency cost the results of this paper could change significantly. 28 This selection has been made by comparing the results report in Appendix Table 3. It is based on a ‘diffuse prior’, where each scenario has equal probability of occurring. WP 03/30 | THREE POLICY OPTIONS FOR CROWN FINANCIAL POLICY 18 The main advantages of the three policies are: • A low weight on distortionary taxation avoids the risk of potentially large losses from a false positive error that could occur if tax smoothing were given high weight; • A low weight on distortionary taxation would confer relatively small losses in two types of situation: o where either or both agency costs and time-consistency are correctly assigned high weights, since in these cases there may be few opportunities for tax smoothing; and o where either or both agency costs and time-consistency are incorrectly assigned low weights, since in these cases a high weight on distortionary taxation would cause significant agency cost and risk of time-inconsistency. The main disadvantages are two-fold. First, in situations where agency cost is insignificant and receives low weight, the opportunity exists to engage fully in tax smoothing without constraint on the level of fungible assets and operating surpluses. False negative errors on distortionary taxation (where distortionary taxation is significant but receives low weight) represent lost opportunities. Second, in addition, a false positive error on agency cost (where agency cost is insignificant but receives high weight) carries the risk of unnecessarily constraining tax smoothing in those cases where distortionary taxation is significant and receives high weight.29 Selection of (L,H,H) as low risk candidate The judgement in this paper is that the policy placing low weight on distortionary taxation and high weights on time-consistency and agency cost should be selected as the low risk candidate. The policy has an advantage over the (L,H,L) policy by ensuring balance between time-consistency and agency cost in their demands for low and high debt targets respectively. The policy is also more definitive in protecting against time-consistency and agency cost than a policy that places low weight on all three issues, (L,L,L). The later policy would carry some risk that the Crown balance sheet would evolve in a random manner. 4.3.1 Medium risk policies A policy that places high weight on all three issues, (H,H,H), is assessed as medium risk. Although this policy is similar in many ways to a policy that places low weight on all objectives, the worst-case scenarios involving poor implementation of tax smoothing and significant distortions to personal portfolios suggest a higher risk assessment. Table 4 - Medium risk policies Policy Options Distortionary taxation Timeconsistency Agency cost 1 7 H L H L H H 29 These cases can be difficult to distinguish. The first is the case of true negative on agency cost and false negative on distortionary taxation. The second is the case of true positive on distortionary taxation and false positive on agency cost. WP 03/30 | THREE POLICY OPTIONS FOR CROWN FINANCIAL POLICY 19 Also assessed as medium risk is the policy that places low weight on distortionary taxation and time-consistency but high weight on agency cost, (L,L,H). Similar to policy (L,H,L), which was assessed as low risk, the (L,L,H) policy is unbalanced across time-consistency and agency cost. The main difference between the two policies is the worst-case scenario. By placing high weight exclusively on agency cost, the (L,L,H) policy would result in a relatively high debt target, thereby increasing the risk of time-inconsistency. The worst-case scenario is that adverse shocks could result in a financial crisis, leading to large reductions in GDP as has been experienced by other countries such as Argentina, Mexico, and the South East Asian countries.30 In contrast, the worst scenario for the (L,H,L) policy is relatively benign, with the high weight on time-consistency potentially leading to excessively low debt target and thereby insufficient discipline on government expenditure. The losses are likely to be much lower than the case of financial crisis associated with the worst scenario for (L,L,H). Selection of (H,H,H) as medium risk candidate The judgement in this paper is that the policy placing high weight on all three issues should be selected as the medium risk candidate. It is a balanced policy that motivates a search for mechanisms to achieve tax-smoothing benefits without jeopardising timeconsistency and agency cost. For example, for distortionary taxation it may be possible to achieve significant benefits through insurance products and possibly derivative instruments (the latter being subject to mitigation of implementation risks). For timeconsistency it may be possible to use the composition of debt to relax constraints. A further advantage is that the (H,H,H) policy would be directly comparable to the low risk candidate (L,H,H), with the only difference being the switch from low to high weight on distortionary taxation. 4.4 High risk policies Three of the eight policy options are assessed as high risk. The three policies have in common that they place high weight on distortionary taxation and low weight on one or both of time-consistency and agency cost. In two of the cases, the policies are assessed as high risk because the high weight on distortionary taxation is not balanced by high weight on agency cost. The absence of a countervailing force implies full engagement in tax smoothing, creating the potential for large losses across a range of situations. Where agency cost is significant but receives a low weight, the strong emphasis on tax smoothing would further exacerbate losses. These situations are assessed as outweighing the advantage that a low weight on agency cost would avoid unnecessarily constraining tax smoothing. This advantage applies only in cases where distortionary taxation is significant and agency cost is insignificant. 30 On the basis of this discussion, a reasonable person could classify the (L,L,H) policy as high risk. For the purposes of this paper the policy has been retained tentatively as medium risk because the number of worst-case scenarios is lower than for the other policies classified as high risk. However, as with all assessments in this paper, future empirical analysis could lead to a change in classification. WP 03/30 | THREE POLICY OPTIONS FOR CROWN FINANCIAL POLICY 26 References Blaug, Mark (1992) “From the received view to the views of Popper.” Chapter 1 in The methodology of economics: Or how economists explain 2nd ed. (Cambridge, UK: Cambridge University Press): 3-26. Bohn, Henning (1990) “Tax smoothing with financial instruments.” American Economic Review 80(5): 1217-30. Bohn, Henning (1995) Optimal Crown Debt: Appendix Report to the New Zealand Treasury by CS First Boston. Bradbury, Simon, Jim Brumby, and David Skilling (1999) “Sovereign Net Worth: An analytical framework.” Treasury Working Paper 99/3 New Zealand Treasury. <http://www.treasury.govt.nz/workingpapers/1999/99-3.asp> Carran, John (2003) “Current Crown balance sheet management and practice.” Internal paper New Zealand Treasury 21 August 2003. Chari, V., Lawrence Christiano and Patrick Kehoe (1994) “Optimal fiscal policy in a business cycle model.” Journal of Political Economy 102(4): 617-52. Christodoulakis, Nicos, David Kemball-Cook and Paul Levine (1993) “The design of economic policy under model uncertainty.” Computational Economics 6: 219-40. Coase, Ronald (1964) The regulated industries [full reference not available]. Coleman, Andrew (1997) “Optimal taxation and the balance sheet.” internal paper, New Zealand Treasury 5 July 1997. Conway, Paul, Aaron Drew, Ben Hunt and Alasdair Scott (1998) “Exchange rate effects and inflation targeting in a small open economy: a stochastic analysis using FPS.” Discussion Paper G99/4 Reserve Bank of New Zealand. < http://www.rbnz.govt.nz/research/discusspapers/dp1999.html - P53_5775> Copeland, Thomas, and Fred Weston (1988) Financial theory and corporate policy 3rd ed. (Addison-Wesley). Cyert, Richard and Morris DeGroot (1987) Bayesian analysis and uncertainty in economic theory (New Jersey: Rowman & Littlefield). Davis, Nick (2001) “Does Crown financial portfolio composition matter?.” Treasury Working Paper 01/34 New Zealand Treasury. < http://www.treasury.govt.nz/workingpapers/2001/01-34.asp> Davis, Nick and Richard Fabling (2002) “Population ageing and the efficiency of fiscal policy in New Zealand.” Treasury Working Paper 02/11 New Zealand Treasury. < http://www.treasury.govt.nz/workingpapers/2002/02-11.asp> WP 03/30 | THREE POLICY OPTIONS FOR CROWN FINANCIAL POLICY 27 Drew, Aaron, and Ben Hunt (1999) “Efficient simple policy rules and the implications of potential output uncertainty.” Discussion Paper G99/5 Reserve Bank of New Zealand. <http://www.rbnz.govt.nz/research/discusspapers/dp1999.html - P53_5775> Fabling, Richard (2002) “Measuring the Crown’s portfolio using a mean-variance approach.” internal paper New Zealand Treasury. Gorringe, Peter (1991) “The methodology of policy advice.” Chapter 10 in Arthur Grimes, Alan Jones, Roger Procter and Grant Scobie (eds.) Economics for policy: Expanding the boundaries: Essays by Peter Gorringe (Wellington, New Zealand: Institute of Policy Studies, Victoria University of Wellington): 230-5. < http://www.treasury.govt.nz/gorringe/bibliography.asp> (refer gp_1991.pdf) Gorringe, Peter (1992) “On Coase and policy.” Chapter 1 in Arthur Grimes, Alan Jones, Roger Procter and Grant Scobie (eds.) Economics for policy: Expanding the boundaries: Essays by Peter Gorringe (Wellington, New Zealand: Institute of Policy Studies, Victoria University of Wellington): 1-12. Gorringe, Peter (1998) “Economic policy choice, decision making under uncertainty and empirical economic research – A framework for considering the issues.” New Zealand Treasury, April <http://www.treasury.govt.nz/gorringe/bibliography.asp> (refer gp_1998.pdf) Grimes, Arthur (2001) “Crown Financial Asset Management: Objectives and practice.” Treasury Working Paper 01/12 New Zealand Treasury. <http://www.treasury.govt.nz/workingpapers/2001/01-12.asp> Hansen, Eric (2003) “Objectives, targets and instruments for Crown financial policy” Treasury Working Paper 03/21, New Zealand Treasury. <http://www.treasury.govt.nz/workingpapers/2003/03-21.asp> Hirshleifer, Jack and John G Riley (1992) The analytics of uncertainty and information (Cambridge: Cambridge University Press). Horman, Greg (2002) “Public Debt Management in New Zealand.” New Zealand Debt Management Office, New Zealand Treasury. <http://www.nzdmo.govt.nz/publications/publicdmnz/public-dm-nz.pdf> Huther, Jeff (1998) “An application of portfolio theory to New Zealand’s public sector.” Treasury Working Paper 98/4, New Zealand Treasury. < http://www.treasury.govt.nz/workingpapers/1998/98-4.asp> Jensen, Michael (1986) “Agency costs of free cash flow, corporate finance, and takeovers.” American Economic Review: 323-39. Knight, Frank (1920) Risk, uncertainty and profit (New York: Harper & Row). Leong, Donna (1999) “Debt management – theory and practice.” Treasury Occasional Paper No.10 (HM Treasury: London). WP 03/30 | THREE POLICY OPTIONS FOR CROWN FINANCIAL POLICY 28 Lucas, Robert and Nancy Stokey (1983) “Optimal fiscal and monetary policy in an economy without capital.” Journal of Monetary Economics 12: 55-94. McCulloch, Brian and Jane Frances (2001) “Financing New Zealand Superannuation.” Treasury Working Paper 01/20 New Zealand Treasury. < http://www.treasury.govt.nz/workingpapers/2001/01-20.asp> Missale, Alessandro (1997) “Managing the Public Debt: The optimal taxation approach.” Journal of Economic Surveys 11(3). Missale, Alessandro (1999) Public debt management (Oxford: Oxford University Press). Onatski, Alexei and Noah Williams (2003) “Modeling model uncertainty.” NBER Working Paper 9566 <http://www.nber.org/papers/w9566> Rhodes, George (1994) “A Bayesian theory of economic policy evaluation.” in George F. Rhodes (ed.) Advances in Econometrics: Simulating and analysing industrial structure, Vol. 10 (Greenwich, Conn.: JAI Press Inc):155-70. Silvey, S.D. (1975) Statistical inference (London: Chapman and Hall) Skilling, David (1997) “A framework for Crown risk management.” Internal paper New Zealand Treasury 11 June 1997. WP 03/30 | THREE POLICY OPTIONS FOR CROWN FINANCIAL POLICY 29 Appendix I: Policy making under uncertainty Formal treatments of Bayesian decision theory are available in Cyert and DeGroot (1987), Hirshleifer and Riley (1992), Rhodes (1994), and Silvey (1975). Gorringe (1991, 1998) discusses the implications for policy making under uncertainty. The purpose of this Appendix is to outline the decision framework in a manner directly applicable to the qualitative analysis in this paper. The first part models the simplest case where policy is described by one variable. The second part generalises the model to the case where policy is described by two variables. One dimensional policy Consider the structure described in Section 3.2 where the true model of the economy is either s or i and the policy choice is either H or L. The policy maker does not know which is the true model but assigns probabilities ps and pi such that ps + pi = 1. The payoffs under each policy option are random variables VTP, VFP, VTN, and VFN. Equivalently, in Bayesian terminology, the policy option should be implemented if the “risk” of the policy option is less than the “risk” of the status quo. The ex ante values of policies H and L are denoted vH and vL, respectively, and are given by: vH = psVTP + piVFP (1a) vL = piVTN + psVFN (2a) Equations (1a) and (2a) may be rewritten as: vH = V - piLFP (1b) vL = V - psLFN (2b) where V = psVTP + piVTN LFP = VTN – VFP > 0 LFN = VTP - VFN > 0 V is the weighted-average value of the true positive and true negative outcomes. It is the ex ante value obtainable if the policy maker could learn the true model prior to making the policy decision, thereby being sure of avoiding false positive and false negative errors. LFP and LFN are the losses associated with false positive and false negative errors that occur when policies H and L are inconsistent with the true model. Equation (1b) says that the value of policy H, vH, is the ex ante value with learning (V) less the probability-weighted loss of policy H being “incorrect” (piLFP). Similarly, equation (2b) says that the value of policy L, vL, is the ex ante value with learning (V) less the probability-weighted loss of policy L being “incorrect” (psLFN). WP 03/30 | THREE POLICY OPTIONS FOR CROWN FINANCIAL POLICY 30 Risk neutral policy maker A risk neutral policy maker would prefer policy L to policy H if and only if the expected value of policy L exceeds the expected value of policy H: E[vL] > E[vH] ⇔ p sE[LFN] < piE[LFP], (3) where E[.] is the expectations operator. Equation (3) indicates that a risk neutral policy maker would prefer policy L if and only if the expected loss is lower than for policy H. Hence, minimising expected losses is equivalent to maximising expected value. Risk averse policy maker A risk averse policy maker would consider both expected losses and risk levels. Bayesian decision theory assumes the policy maker has a preference ordering satisfying standard axioms so that there exists a cardinal utility function, U(.). A risk averse policy maker is assumed to maximise expected utility. Policy L would be preferred if and only if E[U(vL)] > E[U(vH)]. As noted in Section 3.2, a practical consideration in public policy is that the analyst is not the decision maker and under public service conventions would not presume to know the decision makers risk preferences. The proposal in this paper is that the policy analyst should present to the policy maker estimates of expected loss and various measures of risk such as variance, skew and kurtosis. The worst-case loss may be presented also. From the information presented the policy maker would choose the policy option that maximises his or her expected utility. Multi-dimensional policy Section 4.1 noted that for Crown financial policy the policy options are multi-dimensional. The following extends the previous analysis to the case where policy is two-dimensional. With policy defined on two dimensions the set of feasible policies available is {(H, H), (H, L), (L, H) (L, L)}. The set of possible state vectors is {(s, s), (s, i), (i, s), (i, i)}. Analogous to the procedure above, the value of policy (H, H) is: vH,H = ps,sVTP,TP + ps,iVTP,FP + pi,sVFP,TP + pi,iVFP,FP = V – {ps,iLTP,FP + pi,sLFP,TP + pi,iLFP,FP} (5) where px,x is the probability of state (x,x) V = ps,sVTP,TP + ps,iVTP,TN + pi,sVTN,TP + pi,iVTN,TN LTP,FP = VTP,TN - VTP,FP LFP,TP = VTN,TP - VFP,TP LFP,FP = VTN,TN - VFP,FP WP 03/30 | THREE POLICY OPTIONS FOR CROWN FINANCIAL POLICY 31 Equation (5) says that the value of policy (H, H) is the ex ante value with learning (V) less the probability-weighted losses of policy (H, H) being incorrect on one dimension and correct on the other (cases LTP,FP and LFP,TP) or being incorrect on both dimensions (LFP,FP). The full set of solutions for all policy options may be written in matrix notation as: v = V1 + p΄L where V is a scalar and v΄ = (vH,H, vH,L, vL,H, vL,L) 1΄= (1,1,1,1) p΄ = (ps,s, ps,i, pi,s, pi,i) 0 LTP,FN LFN,TP LFN,FN L = LTP,FP 0 LFN,FP LFN,TN LFP,TP LFP,FN 0 LTN,FN LFP,FP LFP,TN LTN,FP 0 WP 03/30 | THREE POLICY OPTIONS FOR CROWN FINANCIAL POLICY 32 Appendix II: Qualitative Assessment of Risks This Appendix applies the indicators developed in Section 3.3 to assess qualitatively the risk properties of alternative policy options. The Appendix is in two parts: Part A discusses the key characteristics of the distortionary tax, time-consistency and agency cost objectives in general terms without reference to specific loss functions. Part B applies these general comments to the loss functions relating to the eight policy options available . Part A: General Discussion 1. Distortionary taxation The two possible decision errors in relation to distortionary taxation are: • False positive error: False conclusion that tax smoothing would confer a significant benefit. A decision to target the minimum-variance portfolio (and possibly build a positive CNW buffer and place an upper bound on total gross debt) that in fact confer no significant benefit relative to the absence of such targets; and • False negative error: False conclusion that tax smoothing is an insignificant issue. A decision to adopt no specific risk/return target (or other bounds) that in fact would confer a significant benefit had they been adopted. Implementation risk Implementation risk could be substantive under the case where tax smoothing is significant. In particular, uncertainty surrounds the risk/return properties of any particular asset or liability, including those on the Crown balance sheet. Information revelation Apart from implementation risks, the probability and timing that a false positive or false negative error would be revealed as a mistake is similar across the two cases. Consider a false positive that tax smoothing would confer significant benefit. Although a stable tax rate would be observed over time, the supposed welfare gains are unobservable. In particular, the difficulties of distinguishing econometrically the tax effects from other influences on economic performance suggests that ex post analysis would be unlikely to yield substantively new and more powerful information than available in the literature currently. (The main information that would come available following adoption of a tax smoothing policy would be the difficulties or otherwise of successful implementation). Similar comments apply in the case of a false negative that the benefits of tax smoothing would be insignificant. WP 03/30 | THREE POLICY OPTIONS FOR CROWN FINANCIAL POLICY 33 Reversibility Assuming a policy mistake did become known, both cases should be low cost to reverse: • in the case where a false positive (in favour of tax smoothing) was revealed, the Crown would be faced with unwinding asset positions that had been built up specifically for smoothing purposes. Provided the unwinding is conducted in an orderly manner, the cost of reversal should be fairly low; and • In the case where a false negative (against tax smoothing) was revealed, the government would have the option of establishing the tax-smoothing regime. The legislative and other institutional arrangements would take some time to work through but would be relatively low cost. Losses would be incurred to the extent that the delay resulted in some or all of the window of opportunity passing by (e.g. with population ageing profile over next few decades). Worst scenario Losses under the worst scenario in the case of a false positive (in favour of tax smoothing) may or may not be larger than in the case of a false negative (against tax smoothing). Under a false positive the worst scenario would be a substantial permanent reduction in asset values accompanied by failure to hedge the Crown balance sheet due to instability in correlations between assets. Under a false negative the worst scenario would be the loss in economic performance of the country due to instability in the tax rate. The loss would be greater to the extent that, for example, population ageing is a one-off change and the delay in implementing tax-smoothing mean the window of opportunity passed by. Dependence on state variables The loss arising from a false positive (in favour of tax smoothing) would be less sensitive to state variables than the loss arising from a false negative (against tax smoothing): • A false positive means, by assumption, either that variation in tax rates has insignificant impact on economic welfare or that implementation is too difficult and costly. • A false negative means, by assumption, that variation in tax rates does have significant impact on economic welfare (over and above the cost of implementation). Therefore the cost of being wrong would depend on the state of the economy and the profile of government expenditure. Vote misperception risk A correct conclusion in favour of tax smoothing faces the problem that the public would tend to look at specific measurable outcomes that may not be closely related to the policy objective. For example, a desirable hedging strategy could lead to an outcome where the market value of Crown financial assets was revised downward by (say) $5 billion as an offset to upward revaluation in the tax asset. Because accounting rules mean that the highly visible headline Operating Balance would include the downward revision in financial asset value but exclude the increase in the tax asset, the public would likely view the outcome as reflecting poor economic management by the government of the day. The hedging benefits are difficult to communicate. In contrast, a correct conclusion against tax smoothing would not face the same issues. WP 03/30 | THREE POLICY OPTIONS FOR CROWN FINANCIAL POLICY 34 A tax smoothing policy shifts the burden of taxation over time. The current situation for New Zealand is that tax smoothing implies accumulating assets in preparation for government expenses associated with population ageing and may also imply a need to build a positive CNW balance (if tax rates and consumption are negatively correlated). These imply bearing the cost of a higher tax rate now in return for lower tax rate (relative to counterfactual) in several decades in the future. 2. Time-consistency of policy The two possible decision errors in relation to the time-consistency of policy are: • False positive error: False conclusion that time-inconsistency is a significant issue. A decision to impose strong upper bound on total net debt (and possibly bounds on other debt variables) that in fact confer no significant benefit relative to the absence of such bounds; and • False negative error: False conclusion that time-consistency is an insignificant issue. A decision to impose no upper bound on total net debt that in fact would confer a significant benefit had they been imposed. Implementation risk Differences in implementation risk are negligible. The main risk facing the decision to impose upper bounds on total net debt relate to judgements about the appropriate level of the bound. The risk is considered fairly small as the incentives on a government to pursue a time-inconsistent policy relate to the broad magnitude of the debt rather than its precise level. Information revelation A false positive that time-inconsistency issues are significant (and hence bounds imposed) has high probability of never being revealed. This is because if timeinconsistency truly is an insignificant issue then inconsistent policies are unlikely to be observed but it may not be clear whether this is due to success of the constraints or lack of underlying issue. In contrast, a false negative that time-consistency is an insignificant issue (and hence no bounds imposed) would mean that future policy inconsistencies would occur with positive probability. At some point as the risk of time-inconsistency increases, the financial markets would be likely to price the risk as an observable interest rate premium on government bonds. Therefore, a false negative is more likely to be revealed than a false positive. WP 03/30 | THREE POLICY OPTIONS FOR CROWN FINANCIAL POLICY 35 Reversibility The reverse conclusion holds in terms of reversibility, since removing debt targets found to be unwarranted may be easier than imposing debt targets in difficult circumstances. Particularly, in the circumstance where the country is caught in a financial crisis it may be very difficult to introduce debt targets. Thus, a false negative would be less easily reversed than a false positive.39 Worst scenario The worst case scenario under the false positive (that time-consistency is significant) relate to the costs of being constrained from structuring the Crown balance sheet to meet other objectives, such as tax smoothing and limiting agency costs. These costs are likely to be moderate. The worst case scenario under the false negative (that time-consistency is insignificant) relate to the costs arising if the country is caught in a financial crisis. The experiences of other countries such as Argentina, Mexico, South East Asia suggests the potential for substantial reductions in GDP. Uncertainty about information revelation Considerable uncertainty exists around the timing of information revelation under the false negative (that time-consistency is insignificant). Ideally, well functioning financial markets would steadily increase risk premia as the risks of time-inconsistency increased. But, equally, given the experiences of other countries, shocks can plunge a country into crisis very quickly and without substantive warning by credit rating agencies and observed risk premia. The false positive (that time-consistency is significant) faces much lower uncertainty. Uncertainty about reversibility Similar conclusions apply to the reversibility of policy. Considerable uncertainty exists whether debt targets would be imposed when the false negative is revealed. Dependence on state variables The losses arising from a false negative (that time-consistency is insignificant) depends on the state of the Crown balance sheet, in particular the government’s future borrowing requirements. Under the false positive (that time-inconsistency is significant) it is less clear how the losses would be affected by the state of the Crown balance sheet or the economy more generally. 39 John Carran has suggested that the opposite conclusion can be justified. The argument is that debt targets may get locked in due to perception and credibility concerns. Thus, the false positive may be quite difficult to reverse. Equally, in the case where a country is caught in financial crisis the government is likely to find it relatively easy to motivate and justify tough debt targets (e.g. NZ reforms of mid-1980s). Thus, the false negative may be relatively easy to reverse. This line of argument, and the implications for policy conclusions, deserves further attention. WP 03/30 | THREE POLICY OPTIONS FOR CROWN FINANCIAL POLICY 42 Var: H The realisation and magnitude of losses depend on occurrence of adverse shocks. Worst: H High losses if the country is caught in a financial crisis, at which time would be difficult to reverse the policy. The experiences of other countries such as Argentina, Mexico, South East Asia suggests the potential for substantial reductions in GDP. LTP,FN,FP = VTP,TP,TN - VTP,FN,FP: The loss when time-consistency is significant but receives low weight and when agency cost is insignificant but receives high weight, while distortionary taxation is significant and correctly receives high weight. Ave: H Losses much higher than previous case as the high weight on agency costs constrains unnecessarily the ability to engage in tax smoothing. Var: H Same as above Worst: H Same as above LFP,FN,TP = VTN,TP,TP – VFP,FN,TP: The loss when distortionary taxation is insignificant but receives high weight and time-consistency is significant but receives low weight, while agency cost is significant and correctly receives high weight. Ave: H The high weight on tax smoothing unnecessarily causes agency costs and also poses implementation risk and distortions to wealth portfolios. With failure to accord high weight to time-consistency, the emphasis on tax smoothing and agency cost also puts timeconsistency at risk. To the extent that policy inconsistencies may occur in future with positive probability, the financial markets would be likely to price the risk as an observable interest rate premium on government bonds. Government would have an incentive to reverse the policy as soon as possible to place high weight on time-consistency. Var: H Same as above Worst: H Same as above LTP,TN,FP = VTP,TN,TN - VTP,TN,FP: The loss when agency cost is insignificant but receives high weight, while distortionary taxation and time-consistency correctly receive high and low weights respectively. Ave: H The high weight on agency cost constrains unnecessarily the ability to engage in tax smoothing. No impact on time-consistency as insignificant in any case. Var: H Same as above Worst: H Same as above LFP,FN,FP = VTN,TP,TN – VFP,FN,FP: The loss when distortionary taxation and agency cost is insignificant but receive high weight and time-consistency is significant but receives low weight. Ave: L The high weights on tax smoothing and agency cost mostly counteract each other, with the exception that both tend to motivate high debt levels. This is limited to the extent that unjustified risk premia occur. To the extent that policy inconsistencies may occur in future with positive probability, the financial markets would be likely to price the risk as an observable interest rate premium on government bonds. Government would have an incentive to reverse the policy as soon as possible. Var: L WP 03/30 | THREE POLICY OPTIONS FOR CROWN FINANCIAL POLICY 43 Worst: H High losses if the country is caught in a financial crisis, at which time would be difficult to reverse the policy. LFP,TN,TP = VTN,TN,TP – VFP,TN,TP: The loss when distortionary taxation is insignificant but receives high weight, while time-consistency and agency costs correctly receive low and high weights respectively. Ave: H The high weight on distortionary taxation unnecessarily causes agency costs and also poses implementation risk and distortion to private portfolios. Var: H Worst: H Poor implementation and significant distortions to private portfolios LFP,TN,FP = VTN,TN,TN – VFP,TN,FP: The loss when distortionary taxation and agency cost is insignificant but receive high weight, while time-consistency is insignificant and correctly receives low weight. Ave: L Var: L Same as for LFP,FN,FP Worst: L 4. Tax smoothing considered unimportant (L,H,H) LFN,TP,TP = VTP,TP,TP – VFN,TP,TP: The loss when distortionary taxation is significant but receives low weight, while time-consistency and agency cost is significant and correctly receive high weight. Ave: L Failure to assign high weight to distortionary taxation causes some losses but extent is limited because time-consistency and agency costs would constrain opportunity for tax smoothing under the correct policy. Var: L Worst: L LFN,TP,FP = VTP,TP,TN – VFN,TP,FP: The loss when distortionary taxation is significant but receives low weight and agency cost is insignificant but receives high weight, while time-consistency is significant and correctly receives high weight. Ave: H The high weight on agency costs constrains unnecessarily the ability to engage in valuable tax smoothing. Var: H Extent of loss depends on occurrence of adverse shocks to economy. Worst: H LFN,FP,TP = VTP,TN,TP – VFN,FP,TP: The loss when distortionary taxation is significant but receives low weight and time-consistency is insignificant but receives high weight, while agency cost is significant and correctly receives high weight. Ave: L Insufficient emphasis on tax smoothing, but losses are limited due to constraints from agency costs. The erroneously high weight on time-consistency could cause agency costs to be higher than warranted but impact limited because the high weight on distortionary taxation under the correct policy would also tend to limit debt levels (up to level where unjustified risk premia occur). WP 03/30 | THREE POLICY OPTIONS FOR CROWN FINANCIAL POLICY 44 Var: L Worst: L LFN,FP,FP = VTP,TN,TN – VFN,FP,FP: The loss when distortionary taxation is significant but receives low weight and time-consistency and agency cost is insignificant but receive high weight. Ave: H Var: H Same as for LFN,TP,FP Worst: H LTN,TP,FP = VTN,TP,TN – VTN,TP,FP: The loss when agency cost is insignificant but receives high weight, while distortionary taxation and time-consistency correctly receive low and high weights respectively. Ave: L Too much weight on agency costs implies high debt target, thereby increasing the risk of time-inconsistency. However, risk is minimal provided time-consistency also receives high weight. Var: L Low variance due to cautionary stance on debt level. Worst: L LTN,FP,TP = VTN,TN,TP – VTN,FP,TP: The loss when time-consistency is insignificant but receives high weight, while distortionary taxation and agency cost correctly receive low and high weights respectively. Ave: L Too much weight on time-consistency inhibits the ability to use high debt level as a discipline on government expenditure, thereby increasing agency costs. The loss occurs only to the extent that the upper bound on debt implied by time-consistency would be less than the lower bound on debt implied by agency costs. Var: L Worst: L LTN,FP,FP = VTN,TN,TN – VTN,FP,FP: The loss when time-consistency and agency cost are insignificant but receive high weights, while distortionary taxation is insignificant and correctly receives low weight. Ave: L The high weights on time-consistency and agency cost is unnecessary but cause little damage. They act to constrain the actions of government but in a relatively harmless way because there is no tax smoothing imperative in any case. Var: L Worst: L WP 03/30 | THREE POLICY OPTIONS FOR CROWN FINANCIAL POLICY 45 5. Exclusive focus on tax smoothing (H,L,L) LTP,FN,FN = VTP,TP,TP - VTP,FN,FN: The loss when time-inconsistency and agency cost are significant but receive low weights, while distortionary taxation is significant and correctly receives high weight. Ave: H An over-emphasis on tax smoothing relative to other issues increases agency cost and time-inconsistency risk through leveraging of the balance sheet. Var: M Losses due to time-consistency depend on occurrence of adverse shocks. Worst: H High losses could occur if agency costs do not become apparent until substantial fund of assets accumulated. Policy reversal could be difficult given political self-interest. High losses if the country is caught in a financial crisis, at which time would be difficult to reverse the policy on time-consistency. LTP,FN,TN = VTP,TP,TN - VTP,FN,TN: The loss when time-consistency is significant but receives low weight, while distortionary taxation and agency cost correctly receive high and low weights respectively. Ave: L Tax smoothing tend to imply leveraging of the balance sheet. However, extent of leveraging would be limited to the level where unjustified risk premia occur. To the extent that policy inconsistencies may occur in future with positive probability, the financial markets would be likely to price the risk as an observable interest rate premium on government bonds. Var: M Depends on occurrence of adverse shocks Worst: H High losses if the country is caught in a financial crisis, at which time would be difficult to reverse the policy on time-consistency. LTP,TN,FN = VTP,TN,TP – VTP,TN,FN: The loss when agency cost is significant but receives low weight, while distortionary taxation and time-consistency correctly receive high and low weights respectively. Ave: H Var: M Same assessment as for LTP,FN,FN Worst: H LFP,FN,FN = VTN,TP,TP – VFP,FN,FN: The loss when distortionary taxation is insignificant but receives high weight and time-consistency and agency cost are significant but receive low weights. Ave: H Var: M Same assessment as previous case Worst: H WP 03/30 | THREE POLICY OPTIONS FOR CROWN FINANCIAL POLICY 46 LFP,FN,TN = VTN,TP,TN – VFP,FN,TN: The loss when distortionary taxation is insignificant but receives high weight and time-inconsistency is significant but receives low weight, while agency cost is insignificant and correctly receives low weight. Ave: M Losses from tax smoothing arise from implementation risk and distortions to private portfolios. Tax smoothing also increases the risk of time-inconsistency up to the extent that unjustified risk premia put limit on leveraging of the Crown balance sheet. Var: L Magnitude of losses not highly dependent on state of economy. Worst: H Poor implementation and significant distortions to portfolios. LFP,TN,FN = VTN,TN,TP – VFP,TN,FN: The loss when distortionary taxation is insignificant but receives high weight and agency cost is significant but receive low weight, while time-inconsistency is insignificant and correctly receives low weight. Ave: H High losses due to agency cost and implementation risk and portfolio distortions Var: H Worst: H LFP,TN,TN = VTN,TN,TN – VFP,TN,TN: The loss when distortionary taxation is insignificant but receives high weight, while time-consistency and agency cost are insignificant and correctly receive low weight. Ave: M Var: L Same as LFP,FN,TN Worst: H 6. Exclusive focus on time-consistency (L,H,L) LFN,TP,FN = VTP,TP,TP – VFN,TP,FN: The loss when distortionary taxation and agency cost are significant but receive low weight, while time-consistency is significant and correctly receives high weight. Ave: L Low loss to the extent the structure of the Crown balance sheet would be similar to the case where high countervailing weights would be placed on both distortionary tax and agency cost. Losses occur to the extent that risks could be hedged without having an adverse impact on agency cost. Losses also occur to the extent that high weight on timeconsistency results in lower debt target, causing higher agency cost. Var: L Relative to the correct policy where hedging would be constrained by agency cost issues, the size of losses would not be sensitive to state of the economy. Worst: M The worst case would be where derivative instruments and insurance products would allow substantive hedging without constraint from agency cost. WP 03/30 | THREE POLICY OPTIONS FOR CROWN FINANCIAL POLICY 47 LFN,TP,TN = VTP,TP,TN – VFN,TP,TN: The loss when distortionary taxation is significant but receives low weight, while time-consistency and agency cost correctly receive high and low weights respectively. Ave: H With agency cost insignificant, the failure to place high weight on distortionary taxation is a lost opportunity. Var: H Size of loss depends on occurrence of adverse shocks Worst: H LFN,FP,FN = VTP,TN,TP – VFN,FP,FN: The loss when distortionary taxation and agency cost are significant but receive low weight and time-consistency is insignificant but receives high weight. Ave: L Var: L Same assessment as for LFN,TP,FN Worst: M LTN,TP,FN = VTN,TP,TP - VTN,TP,FN: The loss when agency cost is significant but receive low weight, while distortionary taxation and time-consistency correctly receive low and high weights respectively. Ave: L The low weight on tax smoothing limits the loss from incorrectly placing low weight on agency cost. Losses arise to the extent that time-consistency results in low debt target that causes higher agency cost. Var: L Worst: L LFN,FP,TN = VTP,TN,TN – VFN,FP,TN: The loss when distortionary taxation is significant but receives low weight and time-consistency is insignificant but receives high weight, while agency cost correctly receive low weight. Ave: H Var: H Same as for LFN,TP,TN Worst: H LTN,FP,FN = VTN,TN,TP – VTN,FP,FN: The loss when time-consistency is insignificant but receives high weight and agency cost is significant but receives low weight, while distortionary taxation correctly receives low weight. Ave: L Var: L Same as for LTN,TP,FN Worst: L WP 03/30 | THREE POLICY OPTIONS FOR CROWN FINANCIAL POLICY 48 LTN,FP,TN = VTN,TN,TN – VTN,FP,TN: The loss when time-consistency is insignificant but receives high weight, while distortionary taxation and agency cost correctly receive low weights. Ave: L The excess weighting on time-consistency motivates a low debt target, but at low cost since tax smoothing offers few gains and high debt targets are not needed to disciple government spending. Var: L Worst: L 7. Exclusive focus on agency cost (L,L,H) LFN,FN,TP = VTP,TP,TP – VFN,FN,TP: The loss when distortionary taxation and time-consistency are significant but receive low weights, while agency cost correctly receives high weight. Ave: L Losses to the extent that some tax smoothing could occur without causing significant agency cost. High debt target to discipline government spending increases the risk of time-inconsistency. Var: L Strong fiscal disciplines allow scope to respond to most adverse shocks, e.g. by increasing the tax rate Worst: H High losses if the country is caught in a financial crisis LFN,FN,FP = VTP,TP,TN – VFN,FN,FP: The loss when distortionary taxation and time-consistency are significant but receive low weights and agency cost is insignificant but receives high weight. Ave: H Losses much larger than the previous case because tax smoothing would not be constrained under the correct policy (where agency cost receives low weight). Var: H Magnitude of losses depend on adverse shocks Worst: H LFN,TN,TP = VTP,TN,TP – VFN,TN,TP: The loss when distortionary taxation is significant but receives low weight, while time-consistency and agency cost correctly receive low and high weights respectively. Ave: L Losses to the extent that some tax smoothing could occur without causing significant agency cost. But assume that high weight on agency cost would limit the extent of tax smoothing in the correct policy. Var: L Worst: L LTN,FN,TP = VTN,TP,TP - VTN,FN,TP: The loss when time-consistency is significant but receives low weight, while distortionary taxation and agency cost correctly receive low and high weights respectively. Ave: L High debt target to discipline government spending increases the risk of timeinconsistency. WP 03/30 | THREE POLICY OPTIONS FOR CROWN FINANCIAL POLICY 49 Var: L Strong fiscal disciplines allow scope to respond to most adverse shocks, e.g. by increasing the tax rate Worst: H High losses if the country is caught in a financial crisis LFN,TN,FP = VTP,TN,TN – VFN,TN,FP: The loss when distortionary taxation is significant but receives low weight and agency cost is insignificant but receives high weight, while time-consistency correctly receives low weight. Ave: H Var: H Same as for LFN,FN,FP Worst: H LTN,FN,FP = VTN,TP,TN – VTN,FN,FP: The loss when time-consistency is significant but receives low weight and agency cost is insignificant but receives high weight, while distortionary taxation correctly receives low weight. Ave: L Var: L Same as LTN,FN,TP Worst: H LTN,TN,FP = VTN,TN,TN – VTN,TN,FP: The loss when agency cost is insignificant but receives high weight, while distortionary tax and time-consistency correctly receive low weights. Ave: L The over-weight on agency cost constrains evolution of the balance sheet but at small cost since both tax smoothing and time-consistency are unimportant. Var: L Worst: L 8. All issues considered unimportant (L,L,L) LFN,FN,FN = VTP,TP,TP – VFN,FN,FN: The loss when distortionary tax, time-consistency and agency cost are all significant but all receive low weight. Ave: L Losses to the extent that some tax smoothing could occur without causing significant agency cost. But assume that high weight on agency cost would limit the extent of tax smoothing in the correct policy. There is no reason to expect the debt target would be higher than under the correct policy, therefore time-inconsistency risks would be no higher. Var: L Worst: L LFN,FN,TN = VTP,TP,TN – VFN,FN,TN: The loss when distortionary tax and time-consistency are significant but receive low weight, while agency cost correctly receives low weight. Ave: H With agency cost insignificant, the failure to place high weight on distortionary taxation is a lost opportunity. With no pressure from either tax smoothing or agency cost, a low debt target would be consistent with low risk of time-inconsistency. WP 03/30 | THREE POLICY OPTIONS FOR CROWN FINANCIAL POLICY 50 Var: H Magnitude of loss depends on occurrence of adverse shocks. Worst: H Loss of tax smoothing benefits in face of adverse shocks LFN,TN,FN = VTP,TN,TP – VFN,TN,FN: The loss when distortionary tax and agency cost are significant but receive low weight, while time-consistency correctly receives low weight. Ave: L Var: L Similar to LFN,FN,FN Worst: L LTN,FN,FN = VTN,TP,TP - VTN,FN,FN: The loss when time-consistency and agency cost are significant but receive low weights, while distortionary taxation correctly receives low weight. Ave: L With low weight on tax smoothing there is no pressure to build up fungible assets, implying low agency cost. With low weight on agency cost there is no pressure to maintain high debt target, so risk of time-inconsistency is low. Var: L Worst: L LFN,TN,TN = VTP,TN,TN – VFN,TN,TN: The loss when distortionary taxation is significant but receives low weight, while time-consistency and agency cost correctly receive low weights. Ave: H Var: H Similar to LFN,FN,TN Worst: H LTN,FN,TN = VTN,TP,TN – VTN,FN,TN: The loss when time-consistency is significant but receives low weight, while distortionary taxation and agency cost both correctly receive low weights. Ave: L Var: L Similar to LTN,FN,FN Worst: L WP 03/30 | THREE POLICY OPTIONS FOR CROWN FINANCIAL POLICY 51 LTN,TN,FN = VTN,TN,TP – VTN,TN,FN: The loss when agency cost is significant but receives low weight, while distortionary taxation and time-consistency both correctly receive low weight. Ave: L With low weight on tax smoothing there is no pressure to build up fungible assets, implying low agency cost. With low weight on time-consistency there is no pressure for very low debt targets, implying relatively low agency cost. Var: L Worst: L Summary of results Appendix Table 3 summarises the results above. The table records for each policy option the number of “high” (H) assessments for the average loss, variance of losses, and worst case. This is based on a ‘diffuse prior’, where the probability of each scenario has equal weight. The allocation of policy options to low, medium and high risk categories reported in Section 2 is based on this table. Appendix Table 3 – Number of “High (H)” assessments Policy options 1 H,H,H 2 H,H,L 3 H,L,H 4 L,H,H 5 H,L,L 6 L,H,L 7 L,L,H 8 L,L,L Average loss 2 6 4 2 4 2 2 2 Variance 2 0 5 2 1 2 2 2 Worst case 4 6 6 2 7 2 5 2