Fiscal rule and public investment in Chile

Fuentes S. M., J. Rodrigo; Schmidt-Hebbel, Klaus; Soto, Raimundo Working Paper Fiscal rule and public investment in Chile IDB Working Paper Series, No. IDB-WP-1189 Provided in Cooperation with: Inter-American Development Bank (IDB), Washington, DC Suggested Citation: Fuentes S. M., J. Rodrigo; Schmidt-Hebbel, Klaus; Soto, Raimundo (2021) : Fiscal rule and public investment in Chile, IDB Working Paper Series, No. IDB-WP-1189, Inter-American Development Bank (IDB), Washington, DC, https://doi.org/10.18235/0003105 This Version is available at: https://hdl.handle.net/10419/237477 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-nc-nd/3.0/igo/legalcode Fiscal Rule and Public Investment in Chile J. Rodrigo Fuentes Klaus Schmidt-Hebbel Raimundo Soto IDB WORKING PAPER SERIES Nº IDB-WP-1189 March 2021 Department of Research and Chief Economist Inter-American Development Bank March 2021 Fiscal Rule and Public Investment in Chile J. Rodrigo Fuentes* Klaus Schmidt-Hebbel** Raimundo Soto* * Pontificia Universidad Católica de Chile ** CIES, Universidad del Desarrollo Cataloging-in-Publication data provided by the Inter-American Development Bank Felipe Herrera Library Fuentes, J. Rodrigo. Fiscal rule and public investment in Chile / J. Rodrigo Fuentes, Klaus Schmidt-Hebbel, Raimundo Soto. p. cm. — (IDB Working Paper Series ; 1189) Includes bibliographic references. 1. Fiscal policy-Chile-Econometric models. 2. Public investments-Chile-Econometric models. 3. Debts, Public-Chile-Econometric models. 4. Budget-Chile-Econometric models. I. Schmidt-Hebbel, Klaus. II. Soto, Raimundo. III. Inter-American Development Bank. Department of Research and Chief Economist. IV. Title. V. Series. IDB-WP-1189 Copyright © Inter-American Development Bank. This work is licensed under a Creative Commons IGO 3.0 Attribution- NonCommercial-NoDerivatives (CC-IGO BY-NC-ND 3.0 IGO) license (http://creativecommons.org/licenses/by-nc-nd/3.0/igo/ legalcode) and may be reproduced with attribution to the IDB and for any non-commercial purpose, as provided below. No derivative work is allowed. Any dispute related to the use of the works of the IDB that cannot be settled amicably shall be submitted to arbitration pursuant to the UNCITRAL rules. The use of the IDB's name for any purpose other than for attribution, and the use of IDB's logo shall be subject to a separate written license agreement between the IDB and the user and is not authorized as part of this CC-IGO license. Following a peer review process, and with previous written consent by the Inter-American Development Bank (IDB), a revised version of this work may also be reproduced in any academic journal, including those indexed by the American Economic Association's EconLit, provided that the IDB is credited and that the author(s) receive no income from the publication. Therefore, the restriction to receive income from such publication shall only extend to the publication's author(s). With regard to such restriction, in case of any inconsistency between the Creative Commons IGO 3.0 Attribution-NonCommercial-NoDerivatives license and these statements, the latter shall prevail. Note that link provided above includes additional terms and conditions of the license. The opinions expressed in this publication are those of the authors and do not necessarily reflect the views of the Inter-American Development Bank, its Board of Directors, or the countries they represent. http://www.iadb.org 2021 1 Abstract* This paper reviews the design and operation of the Chilean fiscal rule in the past 30 years. Using different empirical approaches, we assess its impact on fiscal procyclicality, public debt, and public investment. While there has been substantial progress in building a modern institutional framework for fiscal policy, we find that the rule is incomplete in two dimensions: it lacks an escape clause, and it needs to supplement the budget balance rule with a debt rule. The former is seen in the pervasive inability of the authorities to steer fiscal accounts back to their long-term sustainable path after the rule was breached the rule in 2009. The latter issue is illustrated by the speedy build-up of the public debt as a result of the need to finance fiscal deficits. We do not find, nevertheless, a negative impact of the rule on public investment. We propose reforms to improve on transparency and accountability, as well as to supplement the rule with escape clauses and a debt anchor. JEL classifications: E61, E62, H54, O23 Keywords: Fiscal rule, Public investment, Chile * We gratefully acknowledge useful discussions and comments from Martín Ardanaz, Eduardo Cavallo, Rodrigo Cerda, Alejandro Izquierdo, Sebastián Miller, Teresa Ter-Minassian, Carlos Végh and, especially, Alejandro Vicondoa. Comments received at a Seminar at Universidad de las Américas in Quito improved our work. We also thank Martín Carrasco and Wei Xiong for able research assistance. The usual disclaimer applies. 2 1. Introduction Chile has often been cited as an example of fiscal responsibility in Latin America in the past 40 years (Frankel, 2011a; OECD, 2018). Fiscal responsibility is the result of gradually implementing a modern institutional framework to manage fiscal policy, starting with the 1975 Organic Law of State Financial Administration and the Copper Stabilization Fund of 1985. Chile formally adopted a structural, budget-balance fiscal rule in 2001, enacted a Fiscal Responsibility Law in 2006— including the setup of two sovereign wealth funds—and established an Independent Fiscal Council in 2019. Fiscal rules—in the form of quantitative targets for the government’s budget balance, debt, expenditure, and/or revenue—are a major building block of frontier fiscal institutions. Reasons to adopt a fiscal rule in Chile included strengthening fiscal solvency and sustainability (i.e., attaining sustainable levels of government deficits and public debt), contributing to macroeconomic stabilization (i.e., reducing fiscal policy pro-cyclicality), and making fiscal policy design and execution more resilient to government corruption, political interference, and private-sector lobby influence (i.e., strengthening the political economy of fiscal policy decisions and budget management). Another reason for adopting a fiscal rule is to avoid intergenerational inequity, which would otherwise occur if present generations impose on future generations larger net contributions to government financing than what the former contribute today. Note that this could be the case even when the intertemporal budget gap is equal to zero (Arévalo et al., 2019). As we discuss in this paper, intergenerational equity is not satisfied by Chile’s current fiscal rule. During the past two decades, the Chilean rule has shaped the discussion of the government budget by introducing an explicit conceptual framework (a budget balance rule), relatively transparent procedures to compute the structural balance of the government, and a set of measures to evaluate its outcomes. While successful in general terms, in our view the rule is incomplete and shows significant limitations, particularly regarding the absence of an escape clause when breaching the rule and its neglect of the evolution of public debt. In times of severe stress on fiscal finances as a result of the Covid-19 pandemic and the ensuing decline in economic activity, reviewing the Chilean experience with fiscal institutions is timely and enlightening. In this paper we first assess the evolution, compliance, and effectiveness of the Chilean rule from a historical perspective, considering the achievement of the target and debt sustainability. In Section 2 we describe the political economy of adoption and implementation 3 of the rule in the context of Chile’s overall fiscal framework, the modifications introduced by successive governments, and the deleterious effects of excessive optimism in the estimation of the parameters.1 We evaluate the effectiveness of the rule not only in terms of improving the fiscal stance but also in terms of its ability to isolate fiscal management from the fluctuations of the price of copper, Chile’s main export commodity and an important source of fiscal revenue. We also discuss the breaching of the fiscal rule in 2009 and the difficulties in balancing fiscal accounts afterwards, given the absence of a protocol guiding the authorities on how to breach the rule and return to compliance. This section closes with an analysis of the behavior of public investment over the business cycle, comparing the pre and post-fiscal rule periods and identifying whether the rule has had short or long-term impacts on the level of public investment and/or its trajectory. Having described the mechanics of the fiscal rule in Chile, next we focus on determining whether the observed evolution of key fiscal variables is the result of the rule’s implementation or it simply reflects the path followed by other economies in similar conditions. In Section 3 we use synthetic control methods to conduct a counterfactual analysis of public expenditures, debt, and investment in the pre-rule 1990-2001 period and the post-rule 2001-2019 period, to assess whether the implementation of the rule has had significant effects on the way governments manage their outlays. We consider the implementation of the fiscal rule as an intervention or “treatment” and assess its effects by comparing the observed trajectory of the selected variables to what would have been their path in the absence of the fiscal rule. Our empirical analysis ends in Section 4, where we use a dynamic stochastic general equilibrium model to simulate the response of the Chilean economy to a set of exogenous shocks (productivity, world interest rates, and the terms of trade), with and without the fiscal rule in operation. The shocks are chosen and calibrated to represent the conditions prevalent in Chile around the time of the major breach of the fiscal rule in 2009. The model aims at disentangling the transmission channels of such shocks and highlight their likely effect on public investment and public debt. Our previous analyses contribute to the policy debate with suggestions to improve and/or complement the current fiscal rule, in terms of both its design and its operation. We identify several 1 Elbadawi et al. (2015) and Schmidt-Hebbel and Soto (2018) provide empirical evidence on why countries adopt and maintain rule-based fiscal policy regimes. Their results show that strong fiscal initial conditions, political stability, inflation targets, credible exchange regimes, global financial integration, democracy, and political checks and balances raise the likelihood to adopt and maintain fiscal rules. Chile fulfilled these conditions at the time of adoption. 4 issues. First, considering that the Chilean rule is a cyclically-adjusted (or structural) budgetbalance fiscal rule and that the public debt has increased over the last decade, there is a clear need for attaining debt sustainability, adding to the current budget flow rule a rule on the stocks of public debt and sovereign wealth funds. Therefore, a complementary debt rule would improve on fiscal sustainability and intergenerational equity. Second, the rule makes no reference to the composition of public outlays: particularly government consumption versus public investment. By focusing only on the government’s aggregate balance, it opens the door for less-than-optimal management of public investment, an issue that has become prominent in the literature on Latin America (see Ardanaz et al., 2019; Larraín et al., 2019). Third, in its implementation the rule has proven to be over-optimistic regarding government revenues. While key components and parameters of the rule—such as trend output and the long-term price of copper—are estimated by technical committees with independence of the fiscal authority, in fact their revenue forecasts had been systematically above actual revenues. Such over-optimism leads to persistent failures in meeting the target2 and, if not corrected, might potentially undermine its political legitimacy. Fourth, the issue is compounded by the absence of transparent escape clauses. During its first years of operation, the structural balance target was met; however, this has not generally been the case since the rule was breached in 2009. Lacking an escape clause, the rule is suspended de facto under strong adverse shocks such as the Global Final Crisis in 2008-09 and the current Covid-19 pandemic. The unprecedent fiscal expansion in 2020 will require huge future efforts to ensure fiscal sustainability in Chile, which would be strengthened by adopting the changes in the fiscal rule that are advocated in this paper. 2. Chile’s Fiscal Rule, 2001-2019 Modern institutional frameworks for the conduct of fiscal policy aim at reducing principal-agent problems arising between voters and political authorities due to government impatience, lack of representation of future generations, electoral competition, sensitivity to special-interest lobbies, corruption, and use of asymmetric and biased information (von Hagen, 2005). To overcome these distortions and negative externalities, experts suggest adopting a framework based on nine building blocks (Ter-Minassian, 2011; Debrun et al., 2009): a fiscal responsibility law, a multiyear planning horizon, rules for government asset and liability management, sovereign wealth funds, 2 See Frankel (2011b) for international evidence on this issue. 5 requirements on budget accountability and transparency of the government’s financial management, effective budget planning and execution, external control and auditing, a fiscal council, and, last but not least, a fiscal rule. Evidence suggests that effectiveness of a fiscal rule depends crucially on effective operation of the other building blocks. 2.1 Chile’s Institutional Framework for Fiscal Policy World evidence suggests that the above-mentioned components are not adopted all at once, but gradually and sequentially. Chile is not an exception. In fact, it took the country several decades, as reflected by the timeline depicted in Figure 1.3 Chile’s first institutional building block was an embryonic sovereign wealth fund (SWF)—the Copper Revenue Stabilization Fund (CRSF)— adopted in 1985. The next institutional milestone was reached a quarter of a century later: adoption of the fiscal rule by decree of the Ministry of Finance (MoF)—not by national law—in 2001. In order to strengthen operation and credibility of the rule, an original institutional arrangement was added shortly after starting the rule: projections of two unobservable key variables were outsourced to two committees composed by independent experts: the Advisory Committee for Trend GDP (ACTGDP) in 2001 and the Advisory Committee for the Reference Price of Copper (ACRPC) in 2003. In 2006, Chile enacted its Fiscal Responsibility Law, a comprehensive piece of legislation that improves the overall institutional framework and strengthens the links between the fiscal rule, the establishment of two SWFs, and the use of government savings. This law does not impose any specific fiscal rule on the government. It rather requires adopting and implementing a fiscal policy framework aiming at fiscal sustainability, based on a distinction between actual and cyclicallyadjusted government balances. This provides flexibility to governments to define the formula and parameters for the fiscal rule they commit to at the start of their administration. Finally, the law does not impose restrictions on how budget deficits are financed, maintaining MoF discretion on deciding between issuing public debt, using SWF savings, or selling government assets. 3 A detailed description of the eight building blocks of Chile’s fiscal policy framework—other than the fiscal rule— appears in Appendix A. 12 2.3 Breaches and Changes of the Rule Major international and domestic shocks have hit the Chilean economy since the start of the fiscal rule: the favorable 2004-2013 super-cycle of high commodity prices (implying exceptionally high copper prices), the 2008-2009 Global Financial Crisis and Great Recession, and the adverse, partial reversal of high copper prices since 2014. Two important domestic shocks hit the Chilean economy during the last two decades: the large 2010 earthquake and the political crisis beginning in October 2019 combined with the Covid-19 pandemic since March 2020. During the copper-price boom, the fiscal rule operated flawlessly, shielding public finances from political pressures for higher spending. There is generally strong adherence to the rule in good times. However, the opposite is true during bad times, as discussed next. 2.3.1 Downward Drift in the CAB Target Level, 2001-2010 The rule was adopted in 2001 with a numerical target for the CAB defined at 1 percent of GDP, a surplus intended for funding central bank recapitalization and government pension liabilities. This target was maintained until 2007. Facing huge budget surpluses resulting from the unexpected copper price boom, Minister Velasco reduced the CAB target to 0.5 percent of GDP for 2008 and 2009. The downward drift of the CAB target continued in 2010 and 2011. From 2011 to 2019, the CAB target has remained in the range between -1.0 and -1.8 percent of GDP. 2.3.2 Changes to the Rule in 2009-2010 In response to large adverse shocks to government revenue caused by lower copper prices, the world recession, and the ensuing domestic recession, Minister Velasco introduced in 2009 two major changes to the rule. First, he defined a lower (more negative) target for the CAB to accommodate large countercyclical government expenditure in 2009. Second, to gain additional fiscal space, he changed the rule’s definition by widening the scope of cyclical adjustment to other revenue items and he included temporary tax cuts as “non-structural” (i.e., cyclical) revenue, adjusting the CAB measure upward. On advice of the Corbo Commission (Corbo et al., 2011), the latter methodological changes were reversed by Minister Larraín in 2010, restoring the previous CAB definition that limits cyclical adjustment of revenue to GDP and copper price deviations. Minister Larraín also partially reversed the downward trend of the CAB target in 2011-2013. 13 2.3.3 Changes to the Rule in 2014-2017 In response to declining copper prices and a weak domestic economy, Ministers Arenas, Valdés, and Eyzaguirre lowered CAB target levels, despite the significant increase in tax revenue stemming from the 2014-2015 tax reforms. The latter reforms aimed at raising tax revenue by 3.0 percent of GDP and attaining a zero CAB in 2018 but failed in both dimensions. Expansionary fiscal policy led to a gross public debt increase from 12.7 percent of GDP in 2013 to 25.6 percent of GDP in 2018. 2.3.4 Changes to the Rule in 2019 In 2018 Minister Larraín adopted a fiscal consolidation program, aimed at raising very gradually the CAB target from -2.0 percent of GDP in 2017 to -1.0 percent of GDP in 2021, consistent with stabilizing gross public debt at 26 percent of GDP in 2021. This path of fiscal consolidation was carried out during an 18-month period, through September 2019. Then, as a result of the Chilean domestic crisis that erupted in October 2019, Minister Briones obtained congressional support for an expansionary fiscal program for 2020 and beyond, which implied a significant departure from the previous consolidation. The Covid-19 pandemic and ensuing deep recession in 2020 led the government to implement an additional, very large countercyclical fiscal expansion. Public deficits and debt levels are projected to increase significantly in 2020 and subsequent years. This implies a full reversal of the stance of fiscal policy implemented until September 2019, involving major changes in CAB targets. These modifications to the rule could be justified on the grounds of the imperfections in the rule’s design and/or the perception of its inability to deal with unexpectedly large shocks. Deviations from previous policy could thus be considered second-best changes to the rule. Alternatively, discrete changes in the rule’s definition and parameters, such as those introduced in 2009-2010, can be regarded as unusual breaches of the rule, either in the form of violating the fundamental objectives of the rule or as a failure of the administration to commit to the CAB target. In Section 5 we provide a thorough analysis of this issue, explore the overseeing role given recently to the IFC, and propose reforms aimed at correcting the current limitations of the rule, in order to institutionalize fiscal expansions in response to adverse shocks and strengthen their reversal in subsequent years. 14 3. Forecasts of Key Macroeconomic Variables for the Budget Ex ante projections of trend GDP growth were close to ex post estimations (as of 2019) of trend growth only during the first four years (2003-2006), as seen in Figure 3. Since 2007 trend growth projections have been systematically above their ex post estimations. Average ex ante projected trend growth exceeded ex post estimated trend growth by 0.7 percentage points (pp) per year during 2003-2019. This forecast bias is probably due to the combination of several factors: an unanticipated decline in trend GDP growth since the late 1990s, a large weight of past GDP and production-factor data in future trend GDP forecasts, and limitations of the Solow growth model as the only methodology that is applied in generating forecasts. Figure 3. Ex Ante Projected Trend GDP Growth and Ex Post Estimated Trend GDP Growth, 1991-2019 Sources and notes: See Appendix Table A1. 15 Figure 4. Ex-Ante Projected GDP Growth and Actual GDP Growth, 1990-2019 Sources and notes: See Appendix Table A1. Ex ante projections of actual GDP growth (made since 2001) were also generally larger than actual GDP growth rates (Figure 4). On average, projected GDP growth exceeded actual GDP growth by 0.8 pp. For each and every year in 2001-2019, the ex ante projected output gap has taken a positive value; i.e., governments have projected systematically weak domestic cyclical conditions for the following year (Figure 5). Ex post estimates or backcasts of the output gap are typically lower than ex ante gap projections. The average annual ex ante projected output gap is 2.15 percent, which is significantly larger than the average annual ex post estimate of 0.05 percent. This forecast bias in the output gap (a direct reflection of the bias in GDP forecasts) leads to government overspending and undersaving. 16 Figure 5. Ex Ante Projected Output Gap and Ex Post Estimated Output Gap, 1990-2019 (%) Sources and notes: See Appendix Table A1. ACRCP copper-price projections appear also to be significantly influenced by lagged prices. This leads to under-estimation of future prices during unexpected price booms (in 2004- 2013) and hence over-saving, and to the opposite result afterwards (Figure 6). 17 Figure 6. Actual Copper Price, Ex Ante Projected Copper Price, and Ex Ante Projected Reference (or Long-Term) Copper Price, 1990-2019 (US$/lb) Sources and notes: See Appendix Table A1. In summary, projections of the two key variables used in the Chilean fiscal rule had been subject to systematic biases, significantly overestimating GDP growth and underestimating copper prices. While these biases operate in opposite directions, two wrongs do not make a right. Fiscal authorities have been unwilling or unable to implement remedial measures, upgrading and updating the forecasting methodologies employed by the ACTGDP and ACRPC, thus jeopardizing credibility of the rule. 4. Fiscal Performance under the Rule 4.1 Actual Fiscal Balance and CAB Fiscal consolidation in Chile started well before the fiscal rule was in place. The actual fiscal balance (as share of GDP) averaged 1.2 percent of GDP in 1990-2000 and fell to an average 0.3 percent of GDP since the start of the rule (2001-19), see Figure 7. The actual fiscal balance exhibits a very large trend decline, from a surplus close to 2 percent of GDP in the early 1990s to a deficit close to 3 percent of GDP in the late 2010s. This large change in Chile’s fiscal stance reflects the 18 3 p.p. reduction in the CAB target since the start of the rule and the forecast errors and biases in GDP and copper price projections. Figure 7. Ex Post Estimated Cyclical Adjusted Balance, Actual Government Balance, and Ex Post Estimated Cyclical Balance Component, 1990-2019 (% of GDP) Sources and notes: See Appendix Table A1. Yet the fiscal rule has stabilized public finances, compared to the counterfactual case of absence of the rule. This is apparent when comparing the ex ante CAB target and the ex ante actual government balance (Figure 7). The former is much more stable than the second, and their difference—the cyclical component of the budget—shows a high variance that reflects the GDP and copper-price cycles. 19 Figure 8. Ex Ante Projected Cyclical Adjusted Balance Target, Ex Ante Projected Actual Government Balance, and Ex Ante Projected Cyclical Balance Component, 2001-2019 (% of GDP) Sources and notes: See Appendix Table A1. The ex post measures of the CAB, the actual balance, and their difference (the ex post cyclical budget component), depicted in Figure 7, differ from the e -ante measures in Figure 8, due to the forecast errors and biases that were noted above. Hence the correlation between then corresponding ex-ante and ex-post measures—of around 0.6—are not very high (Appendix A Table A3). 4.2 Government Revenue and Expenditure Cyclicality The fiscal rule has led to significant cyclical delinking of expenditure growth from revenue growth (Figure 9). With the significant increase in GDP and copper-price volatility since 2001, the standard deviation of revenue growth doubled between 1990-2000 and 2001-2019. However, due to the fiscal rule, the standard deviation of expenditure growth increased only slightly. Since 2001, the standard deviation of revenue growth has been almost four times as large as the standard deviation of expenditure growth (Appendix A Table 2). 20 Figure 9. Government Revenue and Expenditure Growth Rates, 1991-2019 (%) Source: Budget Office (DIPRES), Ministry of Finance of Chile. Note: The values for year 2019 are presented in “Informe de Finanzas Públicas: Primer Semestre 2020.” 4.3 Fiscal Pro and Countercyclicality Chile’s fiscal rule has been rather successful in lowering fiscal procyclicality. While the rule’s design implies government expenditure a-cyclicality, the strong discretionary and expansionary fiscal policies adopted in response to adverse shocks (e.g., in 2009 and 2019) imply that Chile has had de facto countercyclical fiscal management. In Figure 10 we reproduce the evidence on fiscal cyclicality in Chile, following the methodology proposed in Schmidt-Hebbel and Soto (2018). The rolling 10-year correlation between the cyclical components of real government expenditure and real GDP has steadily declined from levels above +0.5 before 2001 to levels close to -0.5 after 2007. This shows a strongly countercyclical expenditure behavior during the last decade, reflecting the combination of an a-cyclical fiscal rule (in design) with strongly countercyclical expenditure deviations from the rule during strong adverse shocks. As a result of the latter, the current fiscal balance changes from a-cyclical until the late 1990s to strongly pro-cyclical since the late 1990s. 21 The rolling 10-year correlation between the cyclical components of public investment and real GDP show a similar pattern. Before 2001 public capital expenditures had been mildly procyclical but this reversed in the 2000s when they became strongly countercyclical. The latter results are confirmed by empirical studies for Chile’s fiscal rule, which show that there has been a reduction of the pro-cyclical bias in fiscal policy, de-linking government expenditure from cyclical shocks and strengthening the role of government saving in buffering cyclical shocks (e.g., Larraín and Parro, 2008; Kumhof and Laxton, 2009; Ter-Minassian, 2011; Schmidt-Hebbel, 2010; OECD, 2010; Frankel 2011a; Schmidt-Hebbel 2012). Figure 10. Ten-Year Rolling Correlations of the Cyclical Components of Real GDP and Government Expenditures, Public Investment and Fiscal Balance in Chile, 1989-2014 Source: Authors’ compilation elaboration based on Schmidt-Hebbel and Soto (2018). 4.4 Fiscal Solvency: Gross Debt, Net Debt, and Sovereign Risk Premia Reflecting the strong fiscal stance and high GDP growth between 1991 and 2007, the ratios of gross debt and net debt (gross debt minus SWF assets) to GDP exhibit a negative trend until 2007- 2008 (Figure 11). Since then, the deteriorating fiscal position is reflected in growing levels of gross 28 characteristics should include variables that can approximate the path of the treated country but should not include variables that anticipate the effects of the intervention. The identification assumption of SCM is that if the synthetic control unit provides a good approximation of the outcome for the treated unit in the pre-treatment period, then any subsequent difference between the treated and control units can be attributed to the effect of the intervention (policy) on the outcome. Therefore, the units in the donor pool must not be affected by the treatment. An additional advantage of SCM is that, unlike other techniques such as the difference-in- differences approach, which measure the once-and-for-all effect of a treatment, the SCM can account for the effects of confounders changing over time, by weighting the control group to better match the treatment group before the intervention. In particular, when treatment and potential control groups do not follow parallel trends and difference-in-differences methods lead to biased estimates. Two key issues are to be taken into account when undertaking SCM exercises. First, that the pre-fiscal rule period be sufficiently long to provide for a consistent estimate of the timevarying effects of the fiscal rule. Ferman et al. (2018) find that the lack of guidance on how to choose the matching variables used in the SCM estimator would be asymptotically irrelevant when the number of pre-treatment periods goes to infinity. Second, that potential donors (countries) are properly suited for the exercise. As noted by Abadie et al. (2010), researchers trying to minimize biases caused by interpolating across regions with very different characteristics may restrict the donor pool to regions with similar characteristics to the region exposed to the event or intervention of interest. More recently, Firpo and Possebom (2018) find that the root mean-squared predictionerror (RMSPE) statistic has good properties with respect to size, power, and robustness when selecting the appropriate specification for the SCM estimator. We use the latter criterion when selecting the donor group. 6.2 Application In what follows, we apply the SCM estimator to fiscal balances, public investment, and the stock of public debt. We collected information for the three variables from the IMF for the world sample and from the MoF for Chile.9 We expressed all data as a share of GDP. We also collected data for additional controls –such as private investment, terms of trade and the dependency ratio—that are 9 Data sources and definitions are described in Appendix B. 29 used when estimating the counterfactual scenario. Finally, business cycles and terms of trade shocks were computed using appropriate time-series filters. Data on fiscal rules were obtained from Schmidt-Hebbel and Soto (2018). Their complete database comprises 152 countries and covers the period 1970-2017, including countries and years with and without fiscal rules. Countries that had implemented fiscal rules at the national level were excluded from the pool of donors as they are treated units. Likewise, in order to have a pre-fiscal rule period sufficiently long to provide for a consistent estimate of the time-varying effects, we excluded countries without data 10 years prior to the implementation of the Chilean fiscal rule in 2001. We also excluded countries without full data 10 years after the start of the rule in Chile, in order to have unbiased counterfactuals. Our final database comprises 73 countries and covers the period 1982-2017. SCM models were selected on the basis of three criteria. First, we aimed at minimizing the RMSPE of the models. Second, we aimed at minimizing the distance between the synthetic and actual values of the treated unit. Third, we checked the ability of the donor pool to replicate the evolution of each fiscal variable in the pre-treatment period. Control variables of each model were selected following the empirical models included in Schmidt-Hebbel and Soto (2018) and, inevitably, on the basis of available data for the countries in the donor pool. The first finding highlighted by the SCM analysis is that prior to the implementation of the fiscal rule in Chile, the synthetic and actual fiscal balance evolved in a similar fashion, although with an important difference in their levels as shown in Figure 14, where the vertical dashed line in 2001 identifies the rule’s start. In fact, the actual data indicate that Chile had a fiscal surplus of around 1.15 percent of GDP in the period 1991-2000, while the SCM projects an average level of 0.71 percent of GDP. Therefore, the fiscal rule implemented in 2001 followed the historic trend by setting its initial CAB target at a 1 percent-of-GDP surplus. In this sense, the rule ought not be seen as a disciplining device for fiscal policy, but rather as a mechanism for sustainable public finances. Fiscal balances after the implementation of the CAB rule were driven mostly by the commodity boom of the 2000s. Copper prices more than doubled in real terms over the previous decade and, subsequently, generated a large, sustained fiscal surplus (3 percent of GDP on average between 2001 and 2008). The synthetic control projection for that period is of a balanced budget 30 (-0.13 percent of GDP), which indicates that the Chilean government chose to run a fiscal surplus and save rather than spend the copper revenues from the commodity boom. The fiscal stance drastically reversed in 2009 when. As a result of the domestic recession and discretionary fiscal expansion, the deficit reached an all-time high level of 4.2 percent of GDP and the fiscal rule was breached. The SCM simulation projects a fiscal imbalance of only -2.5 percent of GDP, thus indicating that the fiscal shock in Chile was far more significant than in the control group. Fiscal deterioration was, in our view, the result of two forces: on one hand, the need to boost the Chilean economy, which was adversely affected by the Global Recession, and, on the other hand, the political cycle brought upon by the presidential election of 2009. The fiscal rule was breached again in 2010, allegedly because of the reconstruction effort following a devastating earthquake in February 2010. Budget figures indicate that the 2009-2010 budgets were heavily tilted towards expanding current expenditures and transfers. Political pressures to soften the fiscal rule led the authorities to request a reform proposal from a group of economists10 that recommended lowering the target to 0.5 percent of GDP but not further. This was implemented in 2008, but the target was further lowered to 0 percent in 2009. While subsequent governments have all declared their willingness to abide by the fiscal rule, this commitment has not materialized. Between 2010 and 2013 the government managed to reduce the cyclically-adjusted structural fiscal deficit significantly, but finances worsened again in 2014 and deteriorated alarmingly in the period 2015-2017. The worsening in the fiscal stance after 2014 is more serious than what would be expected from the experience of the control group: as shown in Figure 14, the SCM prediction of the fiscal stance for Chile is significantly higher—by by 1 to 1.5 percentage points of GDP—than the actual fiscal balance. 10 Engel et al. (2007) proposed lowering the fiscal target on grounds that the Central Bank’s debt had declined to manageable levels, exchange-rate risks had diminished, and the sovereign wealth fund had sufficient resources. Marcel (2013) argues that “In 2009, as a temporary response to the international financial crisis, the target was reduced to zero percent of GDP. That was the first time it was explicitly recognized that the target was being used as an instrument of fiscal policy.” The lowering of the target proved, nevertheless, to be permanent. 31 Figure 14. Actual Fiscal Balance and Simulated Counterfactual Fiscal Balance Note: The simulated counterfactual is obtained using synthetic control methods. We now turn to two other components of fiscal policy, aiming at disentangling the role of the fiscal rule in hampering public investment and increasing public debt, particularly during the periods when the authorities aimed at returning to fiscal equilibrium. Figure 15 shows the behavior of public investment over GDP in Chile and its SCM counterfactual. Note that prior to the implementation of the fiscal rule the difference between actual public investment and the SCM forecast is almost nil: on average, the actual public investment ratio to GDP was 3.1 percent, while its SCM counterpart was 3.2 percent. During the first years of the fiscal rule, the SCM projected and the actual public investment ratios to GDP are virtually identical, indicating that complying with the fiscal rule in non-stress periods had no significant effects on public investment in Chile compared to donor countries. However, 2005 marks an inflection point whereby SCM projections grow vigorously: starting that year, the projected public investment ratio to GDP rises quickly, attaining 6 percent and more since 2007 while actual public investment lags significantly behind (slightly below 4 percent of GDP). The difference is not compensated by private concessions which, at the time, had shrunk to around 0.1 -5 0 5 10 Fiscal Balance (% of GDP) 1990 2000 2010 2020 Actual Fiscal Balance Simulated Fiscal Balance (using SCM) 32 percent of GDP (as shown in Figure 13 above). A timid expansion in public investment came to a halt after the 2009 crisis and the breaching of the fiscal rule. The gap between the counterfactual scenario –as projected using the SCM—and actual investment in Chile widens steadily to reach around three percentage points of GDP in 2014. These differences are quite large when considering that total public investment is just (or only) 4 percent of GDP. This implies that the cumulative effect on the stock of public is very significant. Note that public investment declined steadily between 2009 and 2018. The breaching of the fiscal rule led to a significant expansion in current expenditures and transfers, adopted first in response to the domestic effects of the Global Financial Crisis, and then in response to the destruction of the 2010 earthquake. That is, breaching the fiscal rule in 2009 and non-compliance afterwards have taken mainly the form of expanded current expenditures and not by accumulating public physical capital. On average, the actual public investment ratio declined by 20 percent between 2009 and 2018. Figure 15. Actual Public Investment and Simulated Counterfactual Public Investment Note: The simulated counterfactual is obtained using synthetic control methods. 2 3 4 5 6 7 8 9 10 Public Investment (% of GDP) 1990 2000 2010 2020 Actual Public Investment Simulated Public Investment (using SCM) 33 Figure 16 shows evidence of the effects of the 2009-2010 crisis and its aftermath on public debt. We focus on gross public debt, thereby neglecting changes in sovereign wealth funds. Before discussing the evolution under the fiscal rule, it is worth noting that the counterfactual SCM projection reproduces the significant decline in public debt observed in the 1990s. Chile’s public debt stood at a high level of 67 percent of GDP in 1991. Rapid subsequent growth and fiscal prudence reduced the debt to close to 20 percent of GDP at the end of the 1990s, which is the level exhibited by the SCM projection for Chile. From 1995 onwards, both actual and SCM projected levels are very similar. Figure 16. Actual Public Debt and Simulated Counterfactual Public Debt Note: The simulated counterfactual is obtained using synthetic control methods. After implementation of the fiscal rule in 2001 and until 2007, the public debt declined even further, which is coherent with complying with the CAB target of a 1 percent-of-GDP surplus. By 2008 Chile’s public debt reached an all-time low of 3.9 percent of GDP. The SCM estimation suggests a much higher level of public debt of around 20 percent of GDP for the 1990s, which is consistent with both the initial level of public debt and the SCM projection of a balanced budget, as described in Figure 14 above and discussed in the previous paragraphs. 0 20 40 60 80 Public Debt (as percent of GDP) 1990 2000 2010 2020 Actual Public Debt Simulated Public Debt (using SCM) 34 Note that the SCM counterfactual scenario projects a significant jump in public debt between 2008 and 2010 from 23 percent to 32 percent, and subsequently the debt ratio continues rising toward 35 percent of GDP. The actual data shows a different path: the public debt started to rise in 2007, one year before the projected SCM jump, when it grew by around 1 percent of GDP in 2008 despite the 3.5 percent growth in GDP of that year. It continued expanding during and after the 2009-2010 crises and, while the public debt of Chile was below its counterfactual level as of 2017 (at 24 percent of GDP), the gap was closing very fast. In summary, according to these SCM projections, the fiscal rule explains little of the fiscal behavior observed between 2001 and 2018, regarding the evolution of the fiscal balance, public investment, and public debt. Instead, the rule ought to be seen as an institutionalization of the responsibility that characterized Chile’s fiscal performance before the 2008-2010 crises. The severe shocks experienced in the latter period led to non-compliance with the rule and a prolonged period of significant fiscal imbalances, far more serious than what can be explained by our SCM projections. The latter, in fact, suggest authorities should have kept fiscal balances under control (the projected deficit is well below one percent of GDP in any year after 2010). While the actual fiscal deficit is not targeted in the fiscal rule but the cyclically-adjusted fiscal deficit, it should be acknowledged that the authorities have interpreted the compliance with the fiscal rule with significant latitude since 2009 and that, prior to the Covid-19 pandemic, there has been little effort to achieve the balanced budget of the rule. After breaching the rule, successive administrations have all claimed to be returning to compliance, but on timetables that exceeded by a significant margin their tenure in office. This basically leaves the issue to the following administration, and with “intermediate targets” for the CAB that are far from the zero-deficit target. As a result, the CAB target fluctuated between -1.0 percent and -1.8 percent of GDP from 2011 to 2019. 7. Simulating the Effects of the Chilean Fiscal Rule in a DSGE Model "For men are good in but one way, but bad in many." Aristotle, Nichomachean Ethics, Book II, Chapter 6. Applied to our concerns, the above quote tells us that there is one way to comply with a fiscal rule, but it could be breached in many ways. Therefore, when trying to assess the role of fiscal rules in public investment, it would be difficult to build a counterfactual scenario of an economy that does 35 not comply with a fiscal rule, without resulting in arbitrariness or being limited by the particular assumptions chosen to represent fiscal irresponsibility. Our goal is, consequently, rather modest but more realistic. We build a DSGE model to simulate and compare the effects of different macroeconomic shocks on fiscal variables –in particular, public investment—in two representative economies: one is endowed with a CAB rule while the other is not confined by a fiscal rule but has an intertemporally responsible government. In the no-fiscal rule economy, the government is intertemporally solvent, in that it internalizes the present and future costs of its policies and never defaults on the public debt. In the fiscal rule case, which represents the Chilean CAB case, the government complies with the rule at all times. The DSGE model allows us to track and quantify with relative ease the short-term, dynamic response to shocks of tax revenues, expenditures, transfers, and public debt, as well as to evaluate their effect on private consumption, investment, and the different sources of household income. We study the impact of shocks to productivity, the terms of trade, and foreign interest rates observed at the time of the breach of the Chilean fiscal rule in 2009 and assess whether the CAB rule was instrumental in ameliorating or deepening the effects of such negative shocks, in particular regarding their effect on public investment.11 The model explicitly accounts for all tax sources (mining, consumption, labor, and capital taxes) and main public outlays (such as public investment, transfers, and expenditures). The structure is kept simple to highlight the transmission mechanisms of the shocks as well as the role of taxes, expenditures, and public debt. We assume that the economy produces two internationally tradable goods. Sector 1 produces an exogenous quantity of an exportable good. This aims at replicating Chile’s export structure which is based on natural resources (copper and other minerals) that, in the short run, behaves almost as an endowment sector. We also assume that the exportable good is not consumed domestically, since evidence shows that domestic consumption of copper is negligible. Exports pay a royalty or export tax, thus having a significant impact on fiscal revenue. The export-tax rate is set to match the average contribution of the mining sector to government revenue in Chile. Sector 2, on the other hand, produces a tradeable final good, using private capital, labor, and public capital. Public capital operates as a positive externality provided freely by the government (e.g., 11 Previous work that has analyzed the Chilean fiscal rule include Medina and Soto (2016) and Kumhof and Laxton (2009) but with different objectives than those pursued by this paper. The former paper analyzes how the economy responds to a commodity price shock under different types of rules, while the latter investigates whether automatic stabilizers are welfare improving compared to a balanced budget rule. 36 roads, ports, and other public infrastructure). This tradeable good competes with imports and it can be used for consumption or investment.12 There are two types of consumers: an exogenous share of the population has access to capital markets (Ricardian agents) and the rest is restricted to consume their income in every period (hand-to-mouth agents). Private consumption by Ricardian agents—and thereby saving and investment—is determined by a representative agent who maximizes her intertemporal utility function, which depends only on consumption. The agent saves by acquiring capital goods, but their purchase is subject to the payment of an adjustment cost. The private sector does not have access to foreign borrowing or lending (only the government borrows abroad) and, therefore, the private-sector current account is zero). In order to keep the model simple, we assume that households offer inelastically one unit of labor in the market. Wages are nevertheless endogenously determined. Ricardian agents also receive income in the form of profits of firms in sectors 1 and 2 and the returns of capital goods, and pays taxes on consumption and each source of income. They do not receive transfers from the government. The hand-to-mouth representative agent consumes her labor income net of income taxes plus the transfers received from the government. Taxes are set in the calibration to replicate both VAT in Chile as well as average tax rates on labor and capital. The government collects taxes and decides how much to transfer, spend, and invest in public capital. Public investment is subject to an adjustment cost (akin to that of private capital). Transfers and government expenditures have been used by the Chilean government for short-term countercyclical policy, and we take that into consideration. The government can also acquire debt in the international market, paying the risk-free international rate plus a premium that depends proportionally on the ratio of the stock of public debt to GDP (Schmitt-Grohé and Uribe, 2003). In the absence of an objective function for the government, we cannot distinguish between public investment, expenditures, and transfers within the outlays of the government. To deal with this shortcoming we study the relationship between the ratio of government expenditures and transfers to GDP and the output gap. We empirically found that, over the years, the relationship is highly stable and that the fiscal policy has been slightly countercyclical (in line with the descriptive discussion in Section 2.2). This, in turn, implies that public investment has potentially been one 12 The model used here is similar to the one in Chumacero and Fuentes (2006), but with the focus on how the existence of the rule affects macro variables, especially public investment, when the economy is facing a technology or price shock. 37 key adjustment variable allowing the government to comply with the CAB rule (the other key variable is the public debt). When introducing this feature in our DSGE model, we maximize the tension between complying with the fiscal rule and sacrificing public investment and/or accumulating public debt. Finally, we model the Chilean CAB fiscal rule adapting an original version by Rodríguez et al. (2006) as follows: 𝐵𝐵𝑡𝑡∗=𝐵𝐵𝑡𝑡−(𝜏𝜏𝑘𝑘𝑟𝑟𝑡𝑡𝑘𝑘𝑡𝑡+𝜏𝜏𝑙𝑙𝑤𝑤𝑡𝑡𝑙𝑙𝑡𝑡+𝜏𝜏𝑐𝑐𝑐𝑐𝑡𝑡) + �(𝜏𝜏𝑘𝑘𝑟𝑟𝑡𝑡𝑘𝑘𝑡𝑡+𝜏𝜏𝑙𝑙𝑤𝑤𝑡𝑡𝑙𝑙𝑡𝑡+𝜏𝜏𝑐𝑐𝑐𝑐𝑡𝑡)�𝑌𝑌 �𝑡𝑡 𝑌𝑌𝑡𝑡�𝜂𝜂�−𝜏𝜏1𝑝𝑝𝑡𝑡𝑦𝑦1𝑡𝑡 +𝜏𝜏1𝑝𝑝�𝑡𝑡𝑦𝑦1𝑡𝑡 where 𝐵𝐵𝑡𝑡 is the effective balance, 𝑌𝑌𝑡𝑡 correspond to GDP, 𝑌𝑌 �𝑡𝑡 is trend GDP, 𝑝𝑝�𝑡𝑡 represents the longrun price of natural resources, and 𝜂𝜂 is the elasticity of tax collection respect to GDP, which is assumed to be equal to 1. The actual balance is adjusted by net non-mining tax revenues in period t and by tax revenues from private mining-companies in period t plus the revenues from the public company. The DSGE model is calibrated using parameters that characterize the Chilean economy (details on the model and its calibration are presented in Appendix C). We use this framework to simulate the effects on private investment and public debt of three types of shocks faced by the Chilean economy around the GFC of 2008/09 and the subsequent breach of the fiscal rule: (a) a productivity decline of two percent, as estimated by the National Productivity Commission for 2009, (b) a temporary drop in the international price of copper of around 25 percent, as shown in Table 6, and (c) an increase in the international interest rate by one percentage point after the GFC. Chile’s EMBI sovereign spread rose from 0.95 in December 2009 to a maximum of 1.66 in June 2010, before it receded to 1.15 by the end of that year. We simulate the effects of the shocks using impulse-response functions for each shock separately and present the main results in Figures 17 to 19. Shocks take the form of an impulse, that is a once-and-for-all shock given at time 𝑡𝑡= 0. The figures display the simulated evolution of key variables as deviations from their steady-state values. Red lines represent the response of the economy when operating under the CAB fiscal rule; blue lines display the response of the economy in the absence of such rule, but the economy is still ruled by a responsible authority. Since the model has a stationary steady-state, responses eventually dissipate. Consequently, the model best describes the short-run response to essentially transitory shocks. 44 commitment of governments to announce and commit to future fiscal consolidation paths after facing temporary budget shocks. Below, we provide additional reasons to justify the inclusion of a debt component in the Chilean fiscal rule. Second, the budgetary formulation and process should be complemented by continuing assessment and management of the government’s comprehensive balance sheet, according to the principles of modern finance. This should feed into the government’s multi-year investment plan, including all government investment projects and infrastructure concessions programs. Third, we propose a substantial enlargement of IFC’s tasks and responsibilities, with a concomitant expansion of its budget and staff. While fiscal and financial decisions on the balance sheet would remain with the fiscal authority (DIPRES and MoF), we recommend that measurement, analysis, and proposals for improvement of the management of the government’s overall balance sheet should be outsourced to the IFC.14 The IFC should also take over from DIPRES the responsibility of managing the ACTGDP and ACRPC and monitoring of the estimations and projections of trend GDP as well as future copper prices. Finally, it should also assume the responsibility of preparing projections of all additional relevant macroeconomic and sector variables, identified in Section 2, which would be of mandatory use by DIPRES for preparation of its annual and multi-annual budgets and its medium-term fiscal projections, based on the fiscal rule. 8.3 Fiscal Rule As mentioned, several studies provide evidence that adoption of Chile’s fiscal rule (and the complementary fiscal framework) has had positive effects on the conduct of fiscal policy, its credibility, and its effectiveness. Fiscal solvency and sustainability have been strengthened by the rule, at least under normal and favorable macroeconomic conditions.15 The pro-cyclical bias of fiscal policy is avoided by the a-cyclical rule, and intergenerational equity has been strengthened by the funding requirement of the Pension Reserve Fund. 14 The IFC should be entrusted with independence to carry out assessments and studies aimed at understanding and improving the fiscal stance, the sustainability and effectiveness of fiscal policy, and the overall effects of fiscal policy on macroeconomic performance, and initiative to present recommendations on fiscal policy reform, including institutional changes, budgetary formulation and execution, fiscal rule design and implementation, management of the government’s comprehensive balance sheet. 15 It is important to acknowledge that it is difficult to separate purely technical motives from political motives for breaching the rule. 45 There are two main weaknesses that we identify in the design of the actual rule. The first is the lack of a debt anchor, which implies that there is no feedback from the stock of net public debt to government expenditure. It implies that under the current CAB rule, even if strictly enforced or complied with, public net debt could exhibit a drift to any value or, at least, exhibit long periods of drift towards lower values (2002-2008) or higher values (from 2008 to the present), without a self-correcting feedback mechanism from spending to debt. Furthermore, should the government for any reason decide to permanently reduce the level of public debt (e.g., to lower debt service), it would have to change the target of the CAB rule in order to achieve a sustained structural surplus. Changing the CAB target usually entails political strife and uncertainty that could be avoided altogether by complementing the current fiscal framework with a debt-ceiling rule. The rule also lacks escape clauses—i.e., it does not foresee exceptional conditions caused by severe shocks which could require formal temporary suspension of or deviation from the rule. It also does not provide for a protocol to return to the CAB target once the rule has been breached. The decision to breach the rule is left exclusively to the fiscal authority without having to formally identify the conditions that justify the decision, the magnitude of the breach, the instruments involved, and the expected path to return to compliance. The absence of a clear protocol and the fact that the magnitude of the breach is arbitrarily set by the MoF without an independent assessment by the IFC opens the door to political interference and suspicion, as in 2009 when the breaching of the rule had strong support from economists but also incumbent politicians as a means to win the presidential election held that year. Also, there is the key issue of the measures and instruments involved in the breaching of the rule. Since breaching the rule is a transitory phenomenon, instruments and measures ought to be transitory as well. In 2009, a significant fraction of the countercyclical expansion in expenditures was in the form of permanently hiring new public employees. While in principle these additional public workers were hired under fixed-term appointments, in practice their contracts became permanent. The combination of the rule’s a-cyclical nature and its lack of escape clauses has led governments facing major adverse shocks to lower their CAB targets in order to implement countercyclical fiscal policies (in 2009-10 and in 2019-20). This second-best response to a double design failure can be considered a strength when it leads responsible governments to enact 46 countercyclical policy in response to strongly adverse shocks. However, it should be considered a weakness if used by an irresponsible government to adopt an expansionary fiscal policy in the absence of adverse shocks (as in 2014-17). Given that fiscal rules are in place to inhibit irresponsible fiscal policies, we think it would be better to change the current fiscal rule. Moreover, the lack of a debt anchor implies that bygones are treated as bygones: high deficits and debt accumulation during certain periods are not corrected by subsequent efforts to reduce deficits and debt levels. The management of the fiscal rule also shows strengths and weaknesses. Under favorable macroeconomic conditions, the rule has been strongly enforced, enabling the government to accumulate significant savings and avoid political pressures for increased spending. A feature that strengthens the political economy of the rule is the existence of independent committees in charge of forecasting key macroeconomic variables (trend GDP and copper prices). However, CAB targets have exhibited a drift trend toward lower values since 2008. As a result of the latter, and in combination with strong adverse shocks, government deficit and public net debt ratios to GDP exhibit an alarming deterioration. It is often criticized that projections are excessively influenced by recent observations, generating systematic biases in projections of longterm copper prices and of trend GDP, and hence in output gap, cyclically-adjusted revenue, the CAB, and government spending. Despite criticism, the fiscal authority has been reluctant to change methodologies and improve forecasts. Implementation of the fiscal rule and its application to the budget is complex, requiring a combination of assumptions on macroeconomic and sector variables, parameters, and some ad hoc decisions. While significant progress has been made in documenting calculations and projections of cyclical adjustments to government revenue and the budget, it is still not possible for outside observers to replicate DIPRES calculations and projections of the budget or to re-compute the mapping from key and secondary macroeconomic assumptions, and key parameters and elasticities. This lack of full transparency imposes credibility costs in applying the rule. Finally, a major weakness in the implementation of the rule was its breach in 2009, when the MoF and DIPRES introduced major changes in the definition of cyclical adjustments and the rule’s parameters to show significantly larger CAB levels than those obtained under the previous methodology, as discussed in Section 2. In our view, these changes constitute a serious breach of the rule, eroding its credibility and representing a severe blow to the institutional framework of 47 fiscal policy. The presence of the IFC since 2019, however, makes a future repetition of this event unlikely. 8.4 Reform Proposals for the Design and Implementation of the Fiscal Rule To overcome the main drawback in rule design, we propose to complement the existing BBR rule for the CAB with a debt rule (DR). While the CAB rule would apply to annual budget formulation and execution, the DR would be reflected in a numerical target for attaining a stationary net public debt (or net total financial liabilities) as a ratio to GDP. The time path defined for achieving the net public debt ratio would determine the path for annual CAB targets required to achieve the target over the medium term. This would ensure feedback from the stock of net public debt (or total net financial liabilities) to CAB and to public spending and budget balance levels, required for consistency with a permanent-income approach to fiscal policy. This extension of the rule would reinforce the intertemporal consistency, credibility, and effectiveness of Chile’s fiscal policy. Furthermore, it would enforce intergenerational fairness by avoiding future generations to be left to make a larger net contribution to the government’s finances than today’s generation, which would otherwise be the result of mounting public debt. In order to institutionalize changes in short-term CAB targets when facing large adverse shocks, we propose the introduction of an option for the MoF to invoke an escape clause that allows for a change in the government’s previous commitment to the future time path of CAB targets, public deficits, and net debt levels.16 The conditions for invoking the escape clause should be pre-specified in a government document and the IFC should provide an explicit assessment if such conditions are satisfied when the government invokes the escape clause. We think that this alternative provides an explicit recognition of the policy discretion required by governments under severely adverse conditions (such as in 2008-10 and 2019-20) but limits its exercise by defining ex ante the conditions under which an escape clause can be invoked, under the scrutiny of the IFC. The escape clause should also consider a protocol for an effective and swift return to fiscal discipline. 16 An alternative to the escape clauses is to replace the current a-cyclical CAB rule by an explicit countercyclical rule, where the CAB target is changed by a fraction of the projected output gap. This has the advantage of a nondiscretionary change but comes at the cost of over-parameterization of the current rule and its dependence on an ex ante defined, inflexible elasticity of CAB with respect to the output gap, bringing all the measurement error of this variable that has been largely discussed in the literature. Therefore, we do not recommend this alternative. 48 Finally, we recommend that DIPRES provide a detailed, step-by-step calculation of the cyclical components of government revenue and balance, both ex-ante and ex-post. 8.5 The Fiscal Rule and Public Investment Chile’s CAB rule does not distinguish between current and capital expenditure by the government. Therefore, public investment is not shielded from adverse shocks that affect the cyclically-adjusted government stance. However, public investment seems to be largely independent of fiscal revenue cycles both before and during the fiscal rule, perhaps reflecting institutional features (e.g., multiyear planning and budget inertia) that our aggregate analysis cannot identify, It is, nevertheless, important to acknowledge that public investment in Chile is low and that, in our counterfactual analysis, it has been shown that during the period of the implementation of the fiscal rule it displayed poorer performance than the synthetic Chile. In the context of a mechanical model where the government complies with the fiscal rule, as seen in Section 4, public investment will not necessarily suffer when the economy faces negative shocks in terms of trade, productivity or real interest rate. Considering the performance of public investment, we do not favor adding specific conditions to the fiscal rule to deal with public investment. This issue should be addressed as part of a plan for development of the economy. Deciding what is the path that the economy will follow in this respect, it comes as a second stage to decide if the investment is carried out directly by the government or through private concessions. Regarding funding, we propose that public investment should be treated consistently as part of the systematic and comprehensive assessment and management of the general government’s balance sheet, including explicit financial assets and liabilities, real assets (general government real estate and equipment, SOEs, infrastructure, land, national parks, among other), implicit liabilities (e.g., future pension payments), and contingent liabilities (e.g., government guarantees for private concessions and particular bank credit programs). This would imply that public investment is evaluated ex ante and implemented as part of the overall portfolio management of public-sector assets and liabilities. As discussed above, this requires strengthening the institutions for efficient social evaluation and political decision-making regarding public investment projects, PPP concessions on public infrastructure, and financial 49 decisions regarding SOE investment projects. These steps will enhance the quality of public investment jointly with an improvement in the management of public finances. 9. Concluding Remarks This paper has assessed the performance of the balanced budget rule applied in Chile since 2001 using different approaches. It pays special attention to how the rule may have affected public investment in different episodes and circumstances. After summarizing the weaknesses and strengths of the rule, the paper includes several proposals regarding how to improve fiscal institutions. The fiscal rule is based on a cyclically-adjusted budget balance that isolates total government outlays management from the price of copper and from business cycles. To comply with the rule, the government saves transitory revenues and dis-saves them during downturns. This optimal intertemporal behavior can be implemented without a fiscal rule, and Chile had been successfully doing so for over a decade prior to enacting the rule. Why, then, did Chilean authorities feel the need to enact a fiscal rule in 2001? The answer has political and macroeconomic management components. On one hand, there was a pressing need to isolate the management of public finance from increasing political pressure. On the other hand, there was a progressive increase in the volatility of the main sources of government revenues (copper revenues and domestic taxation), which made it necessary to tighten spending to comply with long-run fiscal sustainability. During the copper-price boom of the 2000ss, the fiscal rule operated flawlessly. In 2009, facing the international financial crisis, a countercyclical policy was put in place and the government did not meet the target of a CAB equal to zero. We do not consider that the rule is breached or violated when a countercyclical expansion is implemented by the fiscal authority in response to adverse shocks, leading to discrete changes in current and future CAB targets and actual deficit and debt profiles. Such deviations can be considered second-best changes to the rule, considering its limitations in design: its a-cyclical character, the lack of an escape clause, and the lack of a debt anchor. However, we do consider that discrete changes in the rule’s definition and parameters, such as those introduced in 2009-2010, are breaches to the rule and a severe blow to the institutional framework of fiscal policy. Today, with the IFC as the fiscal policy watchdog, such arbitrary violations would be harder to adopt. 50 Public investment does not seem particularly affected by the adoption of fiscal rule. After the implementation of the rule in 2001, however, public investment over GDP has been persistently below the counterfactual benchmark provided by synthetic control methods, up to 2 percentage points of GDP. This may not seem high, but consider that simulated investment in benchmark Chile was above 6 percent of GDP, while effective public investment Chile was around 4 percent. This gap is not overestimated by the exclusion of public-private partnership investment, since the latter was around 0.2 percent of GDP between 2010 and 2018. The secular decline in public investment after 2010 is nevertheless worrying. We further delve into the transmission mechanisms linking fiscal policy and public investment by simulating an artificial economy hit by the shocks observed around the times of the global recession of 2009 and the breaching of the rule using two alternative scenarios—with and without the fiscal rule. We build a DSGE model that mimics the main features of the Chilean economy in terms of fiscal structure (tax rates, revenue sources, expenditures outlays and transfers to the population), consumption structure (with Ricardian and liquidity constrained consumers), production of natural resources and private investment. We include the Chilean budget balance fiscal rule and, as an alternative, a responsible government that does not follow the rule but respects its intertemporal budget restriction at all times (such as Chile before 2001). Our simulations show that, under the CAB rule, public investment does not bear the brunt of an adjustment when facing negative shocks to productivity, the price of copper, and international interest rate shocks of the size observed in 2009/10. We find that when the economy faces a recession, public investment is less affected under the fiscal rule than in its absence. In the rule-based economy the decline in public investment is milder, but the government must sacrifice more transfers to the population. Furthermore, private consumption declines more in the absence of the fiscal rule, reflecting the differential impact of transfers. As expected, without a fiscal rule, the rise in the transfer component of the economy also increases external debt and the interest rate risk premium. The CAB fiscal rule keeps the level of external debt under control. On the basis of our scrutiny of the historical facts and the empirical counterfactual analysis, we propose several reforms to fiscal institutions. There are two main proposals regarding the design of the rule. First, Chile should establish ex ante escape clauses from the rule that would apply under pre-specified conditions. These escape clauses must include the path that the government will follow to return to compliance of the rule. A rule without ex ante escape clauses, 51 without ex post sanctions for violating the rule and without ex-post corrections of deviations from the rule is weak. It will be breached with probability one. Second, the rule is only a budget balance rule—without a complementary debt rule—and it lacks feedback effects from public debt to government expenditure. This feature may lead to drift in government debt, which is inconsistent with the rule’s objective of fiscal sustainability. As to the institutional framework, we propose moving toward longer-term budget planning, from a one-year horizon to a multi-year horizon. This proposal is consistent with the idea of complementing the existing rule with a debt rule: while the CAB rule would apply to annual budget formulation and execution, the DR would be reflected in a numerical target for attaining a stationary net public debt (or net total financial liabilities) ratio to GDP. The budgetary formulation must be accompanied by a continuous assessment of the government balance sheet, which comprises the management of real, financial and contingent assets and liabilities. This will provide a framework for the government's multi-year budget programming and investment plan. In line with these changes, we propose expanding the IFC’s tasks and responsibilities to include jkindependent authority and initiative to study, assess and propose modifications of the fiscal institutions, budgetary formulation, fiscal rule (design and implementation, which includes a more involvement in parameter estimation), balance sheet management and analysis of macroeconomic effects of the fiscal policy. Substantial progress has been made in terms of the institutional framework for fiscal policy in Chile. Until 2010, the gradual approach to institutional building benefited from political consensus regarding the convenience of fiscal discipline and the absence of major macroeconomic shocks. The global recession proved that Chile had a solid fiscal stance but also evidenced that, when breaching the rule, the authorities had no recourse to steer fiscal accounts back to its longterm sustainable path. As a result, public debt mounted. The need for significant changes to the design and operation of the fiscal rule in Chile is nowadays more pressing than ever, since the fiscal stance has deteriorated significantly and is expected to worsen as a result of the Covid-19 pandemic. 52 References Abadie, A. 2020. “Using Synthetic Controls: Feasibility, Data Requirements, and Methodological Aspects.” Forthcoming in Journal of Economic Literature. Abadie, A., A. Diamond and J. Hainmueller. 2010. “Synthetic Control Methods for Comparative Case Studies: Estimating the Effect of California’s Tobacco Control Program.” Journal of the American Statistical Association 105: 493–505. Alesina, A., and R. Perotti. 1996. “Income Distribution, Political Instability, and Investment.” European Economic Review 40(6): 1203-1228. Alesina, A. et al. 1999. “Budget Institutions and Fiscal Performance in Latin America.” Journal of Development Economics 59(2): 253-273. Ardanaz, M., A. Barreix and L.F. Corrales. 2019. “Las Reglas Fiscales en América Latina.” In: A. Barreix and L.F. Corrales, editors. Reglas Fiscales Resilientes en América Latina. Washington, DC, United States: Inter-American Development Bank. Ardanaz, M. et al. 2019. “Safeguarding Public Investment from Budget Cuts Through Fiscal Rule Design.” Discussion Paper IDB-DP-698. Washington, DC, United States: Inter-American Development Bank. Arévalo, P. et al. 2019. “The Intergenerational Dimension of Fiscal Sustainability.” European Commission Discussion Paper 112. Brussels, Belgium: European Commission. Berganza, J.C. 2012. “Fiscal Rules in Latin America: A Survey.” Documentos Ocasionales 1208. Madrid, Spain: Banco de España. Calmfors, L., and S. Wren-Lewis. 2011. “Fiscal What Are Fiscal Councils, and What Do They Do?” VoxEu.org, 21 April 2011. Available at: https://voxeu.org/article/what-are-fiscal-councils-and-what-do-they-do Cerda, R. et al. 2020. “Understanding Domestic Savings: An Empirical Approach.” Applied Economics 52(6): 905-928. Chumacero, R., and J.R. Fuentes. 2006. “Chilean Growth Dynamics.” Economic Modelling 23(2): 197-214. Chumacero, R., J.R. Fuentes and K. Schmidt-Hebbel. 2004. “Chile Free Trade Agreements: How Big is the Deal?” Working Paper 264. Santiago, Chile: Central Bank of Chile. 53 Chumacero, R. et al. 2007. “Understanding Chilean Reforms.” In: E. Fanelli, editor. Understanding Market Reforms in Latin America, E. Fanelli. London, United Kingdom: Palgrave–McMillan. Corbo, V. et al. 2011. “Propuestas para Perfeccionar la Regla Fiscal.” Comité Asesor para el Diseño de una Política Fiscal de Balance Estructural de Segunda Generación para Chile. Santiago, Chile: Dirección de Presupuestos, Ministerio de Hacienda de Chile Available at: https://www.dipres.gob.cl/598/articles-76544_doc_pdf.pdf Debrun, X., and M.S. Kumar. “The Discipline-Enhancing Role of Fiscal Institutions: Theory and Empirical Evidence.” IMF Working Paper 07/171. Washington, DC, United States: International Monetary Fund. Debrun, X., D. Hauner and M.S. Kumar. 2009. “Independent Fiscal Agencies.” Journal of Economic Surveys 23(1): 44–81. Dirección de Presupuestos, Ministerio de Hacienda de Chile (Dipres). 2018. “Informe de Pasivos Contingentes (2018).” Santiago, Chile: Dipres. Dirección de Presupuestos, Ministerio de Hacienda de Chile (Dipres). 2019. “Indicador del Balance Cíclicamente Ajustado: Metodología y Resultados 2018.” Santiago, Chile: Dipres. Dirección de Presupuestos, Ministerio de Hacienda de Chile (Dipres). 2020. “Informe de Finanzas Públicas, Segundo Semestre 2020.” Santiago, Chile: Dipres. Elbadawi, I., K. Schmidt-Hebbel and R. Soto. 2015. “Why Do Countries Have Fiscal Rules?” Revista Economía Chilena 18(3): 28-61. Engel, E., C. Nielson and R. Valdés. 2011. “Chile’s Fiscal Rule as Social Insurance.” Central Bank of Chile Working Paper 627. Santiago, Chile: Central Bank of Chile. Engel, E., M. Marcel and P. Meller. 2007. “Meta de Superávit Estructural: Elementos para su Análisis.” Santiago, Chile: Dirección de Presupuestos, Ministerio de Hacienda de Chile (Dipres). Mimeographed document. Available at: http://www.dipres.gob.cl/572/articles-21746_doc_pdf.pdf Ferman, B., C. Pinto and V. Possebom. 2018. “Cherry Picking with Synthetic Controls.” Journal of Policy Analysis and Management 39(2): 510-32. Fiess, N. 2005. “Chile’s Fiscal Rule.” In: C. Burnside, editor. Fiscal Sustainability in Theory and Practice: A Handbook. Washington, DC, United States: World Bank. 60 other ministries, regional and municipal governments, independent government agencies and public enterprises; high technical capacity for MoF’s budgetary planning and ensuring its consistency with the government’s fiscal policy objectives; efficient legislative approval of the budget, limiting the possibility of legislative amendments that contradict government fiscal policy; and significant capacity in monitoring and obtaining early information on budget and in implementing budgetary corrections of observed deviations. International evidence shows that countries that meet the latter requirements attain better performance in budget execution, stability, and sustainability of fiscal policy (Alesina and Perotti, 1996; Stein et al., 1998; Alesina et al., 1999; Ter-Minassian, 2010). Chile’s legislation broadly satisfies broadly the latter elements. Its strong presidential system empowers the government with budgetary initiative (not shared with congress) and strong control over budget design and negotiation with Congress.17 The executive holds the sole power of tax initiative. Congress lacks line-item veto and is only entitled to vote for or against the Government’s proposed budget bill. Congress also lacks in-house capabilities to assess and evaluate in depth current fiscal policy, projections in the budget bill, and underlying macroeconomic assumptions. However, significant budget negotiations take place between the government represented by the MoF and congress before the budget comes to vote. The latter negotiations tend to focus on budget (i.e., spending) composition rather than on taxation or the overall resource constraint, not least because of the constraints imposed by the fiscal rule on the budget’s overall resource envelope. Budget Planning Horizon Many countries are reforming their budget planning, shifting from one-year to multi-year horizons. This change is aimed at achieving greater transparency, consistency, and intertemporal sustainability of fiscal policy; reducing government short-term bias in public spending decisions; and committing to a path of gradual correction of initially unsustainable fiscal positions. Adopting explicit multi-year budgetary and fiscal targets contributes to additional potential gains in fiscal policy sustainability and credibility, as well as in macroeconomic stability. 17 See Chumacero et al (2007). 61 Chile has in place a one-year budgeting framework. However, the aforementioned 2006 Fiscal Responsibility Law includes two provisions that require governments to adopt a fiscal policy that is, at least in principle, consistent with a horizon that exceeds one year: i) the requirement imposed on each new administration to present its four-year fiscal policy framework and its implications for the cyclically-adjusted government balance, and ii) the requirement of the MoF to submit annually, together with the draft budget law, a medium-term budget projection. As discussed below, Chile’s fiscal rule introduces an additional dimension of multi-year planning, as it requires medium-term projections for GDP and the copper price for next year’s budget calculation. Among the duties of Chile’s Independent Fiscal Council (IFC) is preparation of long-term projections for government accounts and budgets, based on a systematic assessment of fiscal sustainability and a regular evaluation of long-term consequences of fiscal decisions with budgetary implications. Required by the 2006 Law, the AFCFRF regularly commissions an actuarial study of the contingent liabilities derived from the government’s pension subsidies in order to assess if the latter are matched by Pension Reserve Fund assets. Rules for Management of Government Assets and Liabilities While some governments have adopted a consistent and rule-based framework in managing their overall financial assets and liabilities, most have limited such policy to one part of their balance sheet, focusing only on rule-based financial management of some of their assets (in particular, SWF assets) and/or part of their liabilities (in particular, public debt). Chile (like most countries) lacks both a policy and an institution that focuses on consistent management of all government financial assets and liabilities, with explicit financial objectives linked assessments of risk, return, liquidity, or benchmarking. No comprehensive management of the government’s total balance sheet, including financial and non-financial assets, as well as implicit and contingent financial liabilities, has been adopted. Sovereign Wealth Funds Many commodity-exporting countries have adopted SWFs to save their budget surpluses in these funds, usually invested in internationally diversified portfolios. This is observed in countries with 62 and without fiscal rules. In countries with fiscal rules, accumulation into SWFs and withdrawals from them are largely determined by the rules. The two Chilean SWFs were legally established by the 2006 Fiscal Responsibility Law, as discussed above. Requirements on Budget Accountability and Provision of Information on Financial Management Most governments have improved transparency and accountability of their fiscal policy and financial management. This trend responds to the growing political and academic consensus on the democratic and economic efficiency gains of fuller accountability and transparency in fiscal policy. The International Budget Partnership compiles the Open Budget Survey, an international comparative database on partial and aggregate measures of global transparency and accountability of the budgets of central governments in the world. In its 2017 version, Chile ranks in 35th place among 115 countries. In Chile, transparency in the forecasts of the two key variables for the budget—future GDP growth and copper prices—was ensured almost from the start of the fiscal rule, as projections for the two latter variables made by two ad hoc committees were made public (more on them below). However, mapping of the latter projections into actual MoF budget calculations cannot not be easily replicated by analysts. Some improvements in the provision of information regarding calculation of the cyclically-adjusted balance were implemented in the 2010 Report of Public Finances for the Budget Law, which was consistent with the recommendations provided by the Advisory Committee on Fiscal Policy (2010, 2011). However, it is still not fully possible to replicate the MoF calculation of CAB as of today. Regarding SWFs, international heterogeneity in their management and investment accountability and transparency is large as reflected by the annual rankings of SWF published by the Sovereign Wealth Fund Institute. In its 2019 version, both Chilean SWFs rank in first place (together with 13 other SWFs) in transparency, among 88 included funds. Therefore, the current state of transparency and accountability of Chile’s government accounts is moderately high but there is significant room for further improvement, particularly in budgetary transparency. 63 Budget Planning and Execution The responsibility of proposing the resource allocation of government expenditures relies entirely on the MoF but the budget has to be approved annually by Congress. MoF and more specifically the Budget Office (DIPRES) is responsible for preparing the annual budget and enforcing its execution. However, the MoF has a margin of flexibility to manage budget execution during the year. Every year, the MoF reports on the fiscal stance to Congress and presents the budget for the following fiscal year. The budget discussion between the MoF and Congress is subject to the following restrictions: (i) Congress has a mandatory 60-day limit for approving the budget bill. Failure to do so results on automatic approval of the Government’s budget proposal. (ii) The income calculation is made known to Congress but is not voted on. (iii) Tax legislation is understood as having a permanent nature and cannot be altered by the (annual) budget law. (iv) Congress does not have the ability to raise spending or introduce new spending items. It can only approve or lower the spending outlays proposed by the Government. However, in a strict interpretation of the Constitution, Congress cannot cut or lower them when those cuts preclude exercise of Government functions. Congress cannot cut the proposed outlays for spending items allocated to cover commitments derived from permanent laws (pension payments, salaries to tenured public personnel, debt service, etc.). At the start of a new Government administration, the MoF is required to present a fiscal policy plan that defines numerical annual targets of the CAB for the administration’s four-year period. External Control and Auditing External control and auditing of government accounts and budget execution is performed by congress, the government’s general comptroller or auditor, and fiscal councils and committees. In most countries, congress and especially the general comptroller exercise traditional accounting, financial, and managerial control of budget execution and government accounts. More recently, 64 countries are adopting fiscal councils, which are empowered to provide an economic assessment of fiscal policy design and execution, as discussed below. Government accountability of budgetary compliance in Chile is limited to a large extent to oversight and control by the General Comptroller’s Office (Contraloría General de la República). This institution focuses on accounting, procedural, and legal compliance by the government in its implementation of fiscal policy and execution of the budget. This oversight and control function by the General Comptroller is exercised for all levels of government. In this way the Comptroller’s Office contributes decisively to limiting the extent of illegal government actions and government corruption. However, other than the future work by the recently established independent fiscal council, there is almost no substantive ex-ante, real-time or ex-post evaluation of fiscal policy implementation and its compliance with the fiscal rule, of government spending (either mandatory or discretionary), and of the efficiency and effectiveness of government programs and tax collection efforts. There are very few resources spent by Congress, political parties, think tanks or academia in conducting a systematic evaluation of fiscal policy and its results. The only exceptions to this dearth of fiscal policy evaluation are a few academic papers and some exceptional initiatives at the MoF to assess partial aspects of fiscal policy. Fiscal Council Independent fiscal councils are permanent government or congressional agencies established to monitor budget preparation and execution, assess assumptions and projections required for the budget, assess and recommend on fiscal policy and budgetary management, assess the long-term sustainability and optimality of fiscal policy, as well as its macroeconomic and distributional effects, assess and recommend policies on government asset and liability management, recommend budgetary corrections when facing deviations in budget execution, and recommend improvements in budget transparency (Debrun and Kumar, 2007; Debrun et al., 2009; Ter- Minassian, 2011; Calmfors and Wren-Lewis, 2011). Chile’s MoF established by ministerial decree an Advisory Fiscal Council (AFC) in 2013, with a narrow set of functions. They comprise the following: i) assessment and aggregation of projections of trend GDP growth by ACTGDP and the reference copper price by ACRCP; ii) assessment of the detailed calculation of the budget, based on the projection of the cyclical 65 component of government revenue; iii) assessment of changes to the CAB calculation introduced by the MoF; and iv) advice on fiscal policy issues, on request by the MoF. The five non-remunerated council members were selected by the Minister of Finance for overlapping terms and invited to the MoF to 4-7 meetings per year. AFC lacked independent staff and financial resources. However, its experience was useful in defining a broader role for the subsequent Independent Fiscal Council (IFC). Congress enacted Law 21148 in 2019 that established Chile’s IFC. This institution started in April 2019, replacing the previous AFC. Its five remunerated members, who serve overlapping terms, are nominated by the Government and approved by Congress. The IFC has been endowed with a broader set of tasks, taking over the four functions of AFC and adding several more. Its nine functions are: i) evaluation and monitoring of DIPRES calculations of cyclical adjustments of government revenue; ii) monitoring and evaluation of the projections prepared by ACTGDP; iii) submission of proposals of methodological changes to the calculation of the CAB; iv) expression of opinions regarding deviations in government compliance with the CAB target and presentation of proposals for amending such deviations; v) evaluation of the medium and long-term sustainability of public finances and publication of its findings; vi) advise of the MoF on fiscal issues related to the Council’s functions, at the request of the Minister; vii) preparation of research and analysis of fiscal issues related to the Council’s functions and their publication in reports delivered to Congress; viii) outsourcing of studies and reports on issues related to the Council’s functions; and ix) proposal of substitute members to fill possible vacancies of the advisory committees ACTGDP and ACRPC. While Chile’s new IFC has a broader and more significant role than its predecessor, the AFC, its tasks are still relatively modest in comparison to large and influential fiscal councils or congressional budget offices in other countries. In addition, the IFC’s annual budget and its staffing are proportionately small. 66 Table A1. Variable Definitions and Sources Variable Definition Source Ex ante Projected Reference (or Long-Term) Copper Price Data for year 𝑡𝑡 are projected in year 𝑡𝑡−1 by the Advisory Committee of Long-Term Copper Price. Acta del Comité Consultivo del Precio de Referencia del Cobre Ex ante Projected Copper Price Data for year 𝑡𝑡 are projected in year 𝑡𝑡−1 by the Budget Office (DIPRES), Ministry of Finance of Chile Informe de Finanzas Públicas. Proyecto de Ley de Presupuestos del Sector Público Ex ante Projected Trend GDP Growth Data for year 𝑡𝑡 are projected in year 𝑡𝑡−1 by the Advisory Committee of Trend GDP Acta del Comité Consultivo del PIB Tendencial (various issues) Ex ante Projected GDP Growth Data for year 𝑡𝑡 are projected in year 𝑡𝑡−1 by the Budget Office (DIPRES), Ministry of Finance of Chile Informe de Finanzas Públicas Proyecto de Ley de Presupuestos del Sector Público Ex ante Projected Output Gap*, ** Data for year 𝑡𝑡 are projected in year 𝑡𝑡−1 by the Budget Office (DIPRES), Ministry of Finance of Chile Informe de Finanzas Públicas Proyecto de Ley de Presupuestos del Sector Público Ex ante Projected Actual Government Balance Defined as the difference between government revenue and expenditure, expressed as a ratio of GDP and announced by Budget Office (DIPRES), Ministry of Finance of Chile Actualizaciones de Proyecciones Fiscales 2019-2024, December 2019 Ex ante Projected Cyclical Adjusted Balance Target Defined as the difference between the Ex ante Projected Actual Government Balance and the Ex-ante Projected Cyclical Adjusted Balance Target, expressed as a ratio of GDP and announced by Budget Office (DIPRES), Ministry of Finance of Chile Informe de Finanzas Públicas Proyecto de Ley de Presupuestos del Sector Público Actual GDP Growth Data for year t are the effective value of the GDP growth at year t. Actualizaciones de Proyecciones Fiscales 2019-2024, December 2019 y Informe de Finanzas Públicas: Primer Semestre 2020. Actual Copper Price Data are the actual average value of the copper price at year t Bloomberg Actual Government Balance Defined as government revenue less expenditure, expressed as a ratio of GDP. Series for 1990- 2019 computed by Budget Office (DIPRES), Ministry of Finance of Chile. Actualizaciones de Proyecciones Fiscales 2019-2024, December 2019 Ex Post Estimated Trend GDP Growth Estimate based on historical 1960-2019 data by the Budget Office (DIPRES), Ministry of Finance of Chile Actualizaciones de Proyecciones Fiscales 2019-2024, December 2019 Informe de Finanzas Públicas: Primer Semestre 2020. Ex Post Estimated Output Gap* Estimates based on historical 1960-2019 data, by Budget Office (DIPRES), Ministry of Finance of Chile Actualizaciones de Proyecciones Fiscales 2019-2024, December 2019 and Informe de Finanzas Públicas: Primer Semestre 2020. Ex post Cyclical Adjusted Government Balance Data for year 𝑡𝑡 are calculated in year 𝑡𝑡+ 1 by the Budget Office (DIPRES), Ministry of Finance of Chile. Indicador del Balance Cíclicamente Ajustado. Ex post Estimated Cyclical Balance Component Computed as the difference between the Actual Government Balance and the Ex post Estimated Cyclically Adjusted Budget Balance, expressed as a ratio of GDP. Informe de Finanzas Públicas: Primer Semestre 2020 Notes: (*) Output gaps are computed as (Trend GDP–Actual GDP)/Trend GDP. (**) Value used for Output Gap for year 2010 corresponds to the review of the Output Gap value by the “Comité Asesor para el diseño de una política fiscal de balance estructural de segunda generación para Chile (2011).” This is because the value of 8.3 percent presented in the 2009 “Informe de Finanzas Públicas Proyecto de Ley de Presupuestos del Sector Público” is not consistent with the value of the Ex ante Projected Trend GDP Growth nor with the value of the Ex ante Projected GDP growth presented in the “Informe de Finanzas Públicas Proyecto de Ley de Presupuestos del Sector Público” for that year. (***) Copper prices are measured in U.S. dollars per pound. 67 Table A2. Descriptive Statistics of Key Macro and Budget Variables, 1990-2019 Variable Period Full sample Pre-fiscal rule Post-fiscal rule Avg. s.d Avg. s.d Avg. s.d Copper Revenue (% of GDP) 1990-2019 2.51 2.12 1.80 1.18 2.95 2.45 Non-Copper Revenue (% of GDP) 1990-2019 19.0 1.23 19.32 0.81 18.82 1.42 Ex ante Projected Trend GDP Growth (%) 2003-2019 4.27 0.83 - - 4.27 0.83 Ex post Estimated Trend GDP Growth (%) 1990-2019 4.66 1.62 6.52 0.96 3.68 0.82 Ex ante Projected GDP Growth (%) 2001-2019 4.53 1.11 - - 4.53 1.11 Actual GDP Growth (%) 1990-2019 4.55 2.76 6.03 3.07 3.69 2.22 Ex ante Projected Output Gap (%) 2001-2019 2.15 1.51 - - 2.15 1.51 Ex post Estimated Output Gap (%) 1990-2019 0.05 2.86 -1.91 2.73 1.18 2.30 Ex ante Projected Copper Price (US$/lb) 2001-2019 2.26 1.01 - - 2.26 1.01 Actual Copper Price (US$/lb) 1990-2019 1.95 1.10 0.99 0.19 2.51 1.01 Ex ante Projected Government Balance (% of GDP) 2001-2019 -0.04 2.33 - - -0.04 2.33 Actual Government Balance (% of GDP) 1990-2019 0.65 2.84 1.23 1.47 0.32 3.39 Ex ante Projected Cyclical Adjusted Government 2001-2019 -0.23 1.15 - - -0.23 1.15 68 Balance (Target) (% of GDP) Ex post Estimated Actual Cyclical Adjusted Government Balance (% of GDP) 2001-2019 -0.37 1.33 - - -0.37 1.33 Ex ante Projected Cyclical Balance Component (% of GDP) 2001-2019 0.19 1.82 - - 0.19 1.82 Ex post Estimated Cyclical Balance Component (% of GDP) 2001-2019 0.69 2.77 - - 0.69 2.77 Government Revenue Growth Rate (%) 1991-2019 5.61 9.75 4.79 5.74 6.04 11.44 Expenditure Growth Rate (%) 1991-2019 6.07 2.88 6.16 2.16 6.03 3.25 Sovereign Premium EMBI (bps) 1999-2019 148.57 43.80 185.00 16.97 144.74 44.19 69 Table A3. Correlations between Committee Forecasts and Actual Variables, 1990-2019 Variables Correlation Coefficient Ex ante Projected Trend GDP Growth Ex post Estimated Trend GDP Growth 0.68 Ex ante Projected GDP Growth Actual GDP Growth 0.53 Ex ante Projected GDP Growth Previous Year GDP Growth 0.61 Ex ante Projected Output Gap Ex post Estimated Output Gap 0.48 Ex ante Projected Copper Price Actual Copper Price 0.81 Ex ante Projected Copper Price Previous Year Copper Price 0.96 Ex ante Projected Long-Term Copper Price Actual Copper Price 0.51 Ex ante Projected Long-Term Copper Price Previous Year Copper Price 0.72 Ex ante Projected Government Balance Actual Government Balance 0.64 Ex ante Projected CAB (Target) Ex post Estimated CAB 0.57 Ex ante Projected Cyclical Balance Component Ex post Estimated Cyclical Balance Component 0.65 Ex ante Projected Output Gap Ex post Projected Copper Price Gap 0.24 Sources and notes: The Ex ante Projected Copper Price Gap is defined as: (Ex ante Projected Long-Term Copper Price–Actual Copper Price)/ Ex ante Projected Long-Term Copper Price. The source is the data presented in Figures 2-12. 76 where 𝑦𝑦�𝑡𝑡 corresponds to the trend output estimated by the National Productivity Commission. The second equation is a closing equation for the model that describe the dynamic of government outlays: 𝑙𝑙𝑙𝑙𝑔𝑔𝑠𝑠𝑡𝑡 =𝑙𝑙𝑙𝑙𝑔𝑔𝑠𝑠𝑡𝑡 � + 0.3(𝑏𝑏𝑡𝑡 �−𝑏𝑏𝑡𝑡) (15) where the variables with ~ are the steady state values. External Sector Given that output of sector 1 is constant over time (𝑦𝑦�1), all foreign shocks come from changes in the terms of trade. Terms of trade follows an AR(1) stochastic process: 𝑙𝑙𝑎𝑎𝑝𝑝𝑡𝑡=�1−𝜌𝜌𝑝𝑝�𝑙𝑙𝑎𝑎 𝑝𝑝� 𝑡𝑡+𝜌𝜌𝑝𝑝𝑙𝑙𝑎𝑎𝑝𝑝𝑡𝑡−1 + 𝜀𝜀𝑝𝑝𝑡𝑡 𝜀𝜀𝑝𝑝𝑡𝑡~𝑁𝑁(0, 𝜎𝜎𝑝𝑝 2) (16) where 𝜀𝜀𝑝𝑝𝑡𝑡 is an i.i.d. normal distributed disturbance and 𝜌𝜌𝑝𝑝 is the autocorrelation coefficient. The current account of the economy is given by 𝐶𝐶𝐴𝐴≡−(𝑏𝑏𝑡𝑡−𝑏𝑏𝑡𝑡−1)=𝑝𝑝𝑡𝑡𝑦𝑦�1+𝑦𝑦2𝑡𝑡 −�𝑐𝑐𝑡𝑡+𝑖𝑖𝑡𝑡+𝜃𝜃 2�𝑖𝑖𝑡𝑡 𝑘𝑘𝑡𝑡−1−𝛿𝛿�2 + 𝑠𝑠𝑡𝑡−𝑁𝑁𝑡𝑡�−𝑅𝑅𝑡𝑡−1𝑏𝑏𝑡𝑡 (17) The Decentralized Equilibrium Here we describe the decentralized equilibrium of this economy, where households and firms take as given the action of the government. The vector of the state variables is given by 𝜈𝜈= {𝑘𝑘,𝑝𝑝,𝑧𝑧,𝑅𝑅}. The competitive equilibrium is defined as a set of allocations rules 𝑐𝑐=𝑐𝑐(𝜈𝜈),𝑘𝑘𝑡𝑡+1 = 𝑘𝑘′(𝜈𝜈),𝑙𝑙=𝑙𝑙(𝜈𝜈) and a set of pricing function 𝑟𝑟=𝑟𝑟(𝜐𝜐),𝑤𝑤=𝑤𝑤(𝜐𝜐), such that: • Households solve the problem (1) subject to (2) and (3), taking as given vector 𝜈𝜈 and the functions 𝑟𝑟(𝜈𝜈),𝑤𝑤(𝜈𝜈). • Firms of the import competition sector solve problem (5), taking as given vector 𝑠𝑠 and the functions 𝑟𝑟(𝜈𝜈),𝑤𝑤(𝜈𝜈). • The economy-wide resource constraint holds in each period and is given by (15). Calibration of the Model Table A3.1 shows the value of the parameters used in the calibration of the model and simulation of the shocks. A time period is defined as a quarter. The tax rates correspond roughly to the average 77 tax rates of the Chilean economy. The tax rate on capital has changed from 0.17 to 0.27 during the period of the operation of the fiscal rule. The stochastic process for technological shock 𝑧𝑧 was estimated using data on TFP from DIPRES. The stochastic process of the terms of trade was estimated using terms of trade for Chile obtained from the World Bank. The parameters of the stochastic process for the interest rate were taken from Chumacero et al. (2004). Table A3.1. Calibration Parameters Preference parameters Steady States values mining sector 𝜸𝜸=1.1; 𝜷𝜷=0.99, 𝝌𝝌=𝟎𝟎.𝟒𝟒 𝑦𝑦1 �= 1; 𝑝𝑝�= 1 Technology parameters Stochastic process 𝜶𝜶= 𝟏𝟏 𝟑𝟑 ; 𝝀𝝀= 0.05 𝜌𝜌 𝑧𝑧 = 0.86; 𝜌𝜌 𝑝𝑝 =𝜌𝜌 𝑅𝑅 = 0.9 𝑨𝑨𝟎𝟎=𝟑𝟑, 𝜹𝜹=𝟎𝟎.𝟎𝟎𝟎𝟎 𝜎𝜎𝑧𝑧= 0.09; 𝜎𝜎𝑝𝑝= 0.32 𝝍𝝍=𝜃𝜃= 0.025 𝜎𝜎𝑅𝑅= 0.001 Taxes External sector 𝝉𝝉𝒍𝒍=𝟎𝟎.𝟏𝟏 ; 𝝉𝝉𝒌𝒌=𝟎𝟎.𝟐𝟐 𝑏𝑏 �= 3; 𝑅𝑅 �= 0.03 𝝉𝝉𝒄𝒄=𝟎𝟎.𝟏𝟏𝟏𝟏 ; 𝝉𝝉𝟏𝟏=𝟎𝟎.𝟎𝟎𝟎𝟎 𝜙𝜙= 1.2