Policy Research Working Paper 8964 On Promoting Fiscal Discipline The Role of Exchange Rate Regimes, Fiscal Rules and Institutions Kady Keita Camelia Turcu Governance Global Practice August 2019 Policy Research Working Paper 8964 Abstract This paper examines how fiscal rules, exchange rate regimes, is not sufficient to promote countercyclical fiscal policy and institutional quality affect the cyclical behavior of fiscal and should be combined with strong institutions. More- policy (how government spending responds to fluctuations over, fiscal rules seem to limit procyclicality, especially in in gross domestic product). The analysis is performed countries with flexible exchange rate regimes rather than on a panel of 153 advanced, emerging, and developing in countries with fixed exchange rates. The analysis also countries over 1993–2015 using local Gaussian-weighted finds that the disciplining effect of fiscal rules depends on ordinary least squares and two-stage least squares estimators. the type of rule. The findings show that the adoption of fiscal rules alone This paper is a product of the Governance Global Practice. It is part of a larger effort by the World Bank to provide open access to its research and make a contribution to development policy discussions around the world. Policy Research Working Papers are also posted on the Web at http://www.worldbank.org/prwp. The authors may be contacted at kkeita2@worldbank.org. or at kady-synthia.keita@univ-orleans.fr or camelia.turcu@univ-orleans.fr. The Policy Research Working Paper Series disseminates the findings of work in progress to encourage the exchange of ideas about development issues. An objective of the series is to get the findings out quickly, even if the presentations are less than fully polished. The papers carry the names of the authors and should be cited accordingly. The findings, interpretations, and conclusions expressed in this paper are entirely those of the authors. They do not necessarily represent the views of the International Bank for Reconstruction and Development/World Bank and its affiliated organizations, or those of the Executive Directors of the World Bank or the governments they represent. Produced by the Research Support Team On Promoting Fiscal Discipline: The Role of Exchange Rate Regimes, Fiscal Rules and Institutions∗ † Kady Keita Camelia Turcu‡ Keywords: Cyclicality of fiscal policy, Exchange rate regimes, Fiscal rules, Institutions, 2SLS. JEL classification: E32, E62, O23. ∗ We thank participants to the Annual Doctoral Meeting on Development 2017, ICEFM 2017, AFSE 2018, INFER Annual Conference 2018 and FIBA 2018 for insightful comments. Special thanks to Carine Meyimdjui, Marin Ferry, Yannick Lucotte, Tobignar´ e Yabr´ e Tapsoba, Alexandru Minea, Peter Claeys, Helena Sanz Morales and Eckhard e, Ren´ Janeba for constructive remarks that helped us improve the paper. All remaining errors are ours. † PhD Candidate University of Orl´ eans, LEO CNRS and Consultant, World Bank. Email: kady- synthia.keita@univ-orleans.fr ‡ eans, LEO CNRS and INFER. Email: camelia.turcu@univ-orleans.fr Professor, University of Orl´ 1 Introduction The global financial crisis and the critical issue of sovereign debt solvency faced by many advanced economies, including the European Union members, have recently raised the question of how coun- tries can pursue macroeconomic stabilization policies while maintaining the growth of public debt on a sustainable path. A mechanical solution that has been proposed in the literature is that countries pursue fiscal stimulus measures during a downturn in the economic cycle (bust) followed by fiscal contraction (and debt reduction) in a period of recovery in the economic cycle (boom). In other words, countries are encouraged to pursue a counter-cyclical fiscal policy. The convenience and usefulness of counter-cyclical policy for debt management and macroeconomic stability has been analyzed, theoretically, by Christiano et al. (2011) and Nakata (2011). They used a stochastic model with sticky prices and found that the optimal policy in such a context is counter-cyclical. However, the literature confirms that fiscal policy is not always counter-cyclical: it remains pro- cyclical1 in many countries, particularly in developing countries (Gavin and Perotti (1997), Tornell and Lane (1999), Alesina and Tabellini (2005), Talvi and Vegh (2005), Diallo (2009), Thornton (2008), Frankel et al. (2013), Bova et al. (2014)). An interesting question that emerges in this framework is how to reduce the pro-cyclicality of fiscal policy and promote more counter-cyclical policies, pledge of fiscal discipline. A large body of the literature addressing the cyclicality of fiscal policy relates it to exchange rate regimes and aims at finding which type of regime (rather fixed exchange rate regimes or more flexible ones) has disciplining effects on fiscal policy. In the recent period, especially in the context of the global financial crisis and of the European sovereign debt crisis (Bergman et al. (2016)), fiscal rules, instruments used to guide fiscal actions, have started to gain more and more interest. Fiscal rules are defined as “agreements designed to mitigate the deficit bias and promote fiscal discipline by constraining budgetary aggregates through numerical limits” (Schaechter et al. (2012)). In 2015, 89 countries had adopted national and supra-national rules to govern their fiscal policy (IMF Fiscal 1 Fiscal policy is considered pro-cyclical when government chooses to increase public spending and reduce taxes during an economic boom, but to reduce spending and increase taxes during an economic recession. 2 Rules Dataset, 2015 and Budina et al. (2012)), compared to only 10 in the 1990s. Despite this increased popularity of fiscal rules, relatively little work has been undertaken on their functioning. We try to fill this gap by examining the effectiveness of different types of fiscal rules in promoting counter-cyclical fiscal policies. Moreover, we focus on the role of institutional quality and try to find whether there is a more optimal solution to promote counter-cyclical fiscal policy (i.e. which deals with the underlying political source of the problem). This analysis is motivated by Frankel et al. (2013), Bergman and Hutchison (2015) and Bergman et al. (2016) which argue that good and strong institutions have beneficial effects on public finance. In addition, we revisit the effect of exchange rate regimes on the cyclicality of fiscal policy and more interestingly, we try to find whether the effect of fiscal rules depends on the exchange rate regime in place.2 In other words, we focus on the effect of rules conditional on the exchange rate regime in place (in line with a Mundell-Fleming approach). To the best of our knowledge, this issue has not yet been addressed in the literature. The contribution of this paper to the literature is threefold. First, it contributes to the emerging literature on the impact of institutions and fiscal rules on the cyclicality of fiscal policy since it em- pirically addresses the issue of their complementarity or substitutability (Bergman and Hutchison (2015), Wyplosz (2005) and Wyplosz (2012)) in reducing pro-cyclicality and promoting fiscal disci- pline. Second, to the best of our knowledge, this is the first paper which examines the disciplining effect of fiscal rules, conditional on the exchange rate regimes. Third, it tackles the heterogeneity of fiscal rules since it examines the disciplining effect of different types of rules (expenditure rules, budget balance rules, debt rules and revenue rules). We use a panel of 153 advanced, emerging and developing countries and employ a two-step econo- metric approach. We first compute the cyclicality coefficients of fiscal policy using the Local Gaussian-Weighted Ordinary Least Squares (LGWOLS): this allows us to obtain time-varying and country-specific coefficients. Second, we estimate how institutional quality, fiscal rules, exchange rate regimes and their interactions affect the predicted cyclicality. In addition to basic panel re- 2 For example, the impact of fiscal rules in a monetary union has been tackled theoretically by Colciago et al. (2008). 3 gression, we use an instrumental variable method, the two-stage Least Squares (2SLS), in order to address potential endogeneity. We show that fiscal rules alone are not enough to promote counter-cyclical fiscal policy and should be combined with better institutions. Moreover, we find that under flexible exchange rate regimes, fiscal rules are more effective in promoting counter-cyclical policies. However, rules seem to be- come counter-productive under more fixed regimes. Pegged exchange rate regimes alone, as well as strong institutions, promote counter-cyclical fiscal policy though. We also find that the disciplinary effect of fiscal rules depends on the type of rule. Specifically, while expenditure rules and budget balance rules seem to favor fiscal pro-cyclicality, debt rules and revenue rules are associated with more budgetary discipline. The rest of the paper is structured as follows. Section 2 provides a brief literature review on linkages of exchange rate regimes, fiscal rules and institutions to fiscal discipline. Section 3 presents our methodology and estimation strategy. Section 4 describes the data and provides a statistical outline. Section 5 presents the estimation results. Section 6 addresses the endogeneity issue while section 7 discusses the robustness. Finally, section 8 concludes and discusses policy implications. 2 Literature review Three main theoretical explanations of fiscal indiscipline emerge from the political-budget cycle models. These models establish a correlation between the electoral calendar and the macroeco- nomic and budgetary forecasts. The first explanation is the “inter-temporal incoherence”. In fact, governments may have an interest in pursuing unexpected expansionary fiscal policy since it would generate inflation that can erode the real value of public debt (Persson and Svensson (1989)). The second explanation is “myopia” among governments and voters. In the context of election uncer- tainty, a government can strategically issue public debt by executing excessive deficits in order to reduce the latitude for its successor (Alesina and Tabellini (1990), Persson and Svensson (1989)). Finally, a third explanation is the “common pool problem” (Tornell and Velasco (1992)) that can 4 arise because of spending decisions that tend to target specific interest groups but whose total costs are socialized. In the vein of the “common pool problem”, the model of Tornell and Lane (1999) explains the pro-cyclicality of fiscal policy by the “voracity effects” which are due to the actions of several politicians to appropriate resources that are common. This is even more true in times of export booms. Indeed, the authors argue that if fiscal policy is decided on a decentralized basis, with many interest groups, the classical inter-temporal smoothing behavior disappears and gives rise to significant expenditure shares in temporary positive income shocks and thus leading to less savings to cope with future negative income shocks. A range of the literature addressing the cyclicality of fiscal policy relates it to exchange rate regimes. This strand focuses on the disciplining effects of alternative exchange rate regimes on fiscal policy. However, it does not provide a clear-cut response. While certain authors give arguments in favor of the conventional wisdom according to which fixed regimes have a disciplining effect on fiscal policy, others support the opposite view underlining that flexible regimes are associated with more discipline. A third group of authors argue that neither fixed, nor flexible exchange rate regimes have a disciplining effect on the conduct of fiscal policy (Gavin and Perotti (1997), Kaminsky et al. (2004)). as and Rose (2001) stand at odds with the conventional view, they find that The findings of Fat´ belonging to a currency union did not procure fiscal discipline. Tornell and Velasco (1995) analyzed the European and Sub-Saharan African countries’ experience and strongly reject the conventional view. Schuknecht (1999) uses a political economy model and shows that governments try to improve their re-election prospects by executing expansionary fiscal policies, but only in countries with fixed exchange rates and adequate reserve levels. For some countries, this raises doubts about the usefulness of fixed exchange rates for stabilizing the macro economy, unless reforms of the institutional framework reduce the scope for election-oriented fiscal expansion. Alberola and Molina (2004) give evidence of a weak relationship between exchange rate regimes and fiscal discipline in emerging markets. They show that fixing the exchange rate regime offsets the disciplinary effects on fiscal policy by easing the constraint on government budget. Duttagupta and Guillermo (2006) 5 found that currency union as fixed regime encourages, unlike flexible regimes, over-spending and free-riding behaviors. The conventional view stipulates that exchange rate-based stabilization induces more discipline than do other anchors. A number of authors give arguments in favor of that conventional wisdom according to which more fixed exchange rate regimes have a disciplining effect on fiscal policy. For example, Canavan and Tommasi (1997) use a theoretical Barro-Gordon model with incomplete information and imperfect monitoring and document that serious stabilizers prefer more visible anchors like the nominal exchange rate. Moreover, Beetsma and Bovenberg (1998) support that monetary unification reduces inflation, taxes and public spending. These disciplining effects of monetary union, which become stronger when the union widens, are likely to raise welfare. Fur- thermore, Canzoneri et al. (2001) provide another theoretical evidence of the conventional view by distinguishing between Ricardian and Non-Ricardian frameworks. They found that under a Ricardian framework, fixed regimes are disciplining in the sense that government respects its inter- temporal fiscal constraint. Empirical studies performed in the same vein document this conventional claim. Ghosh et al. (2010) explain that a fixed regime is disciplining for fiscal policy. They argue that pegged regimes put constraints on the conduct of macroeconomic policy. Under such regimes, the domestic monetary policy follows the anchor country’s monetary policy. They further qualify the pegged regimes as a double-edged sword tool: they are useful for countries lacking institutional credibility and discipline, but, in the same way, they constrain the use of stabilization tools such as the interest rate to offset the macroeconomic shocks that countries could face. Furthermore, Sow (2015) examined the impact of exchange rate regimes on the cyclical properties of fiscal policy for a sample of 118 developing countries over the period 1993-2007. The author found that the magnitude of pro-cyclicality is reduced for countries under pegged regimes compared to those under flexible ones. During the recent decades, many solutions, including fiscal rules have been proposed in the literature to reduce pro-cyclicality of fiscal policy and more generally to promote fiscal discipline. Debrun et al. (2008), for example, identify four main categories of solutions: “(i) policy makers can be 6 held more accountable for their actions (Corbacho and Schwartz (2007)); (ii) improve procedures for the preparation, approval and implementation of annual budget laws (Von Hagen and Harden (1995)); (iii) delegate policy or aspects of fiscal policy to institutions that are isolated from short- term political pressures (Wyplosz (2005)); and (iv) limiting the discretion of budgetary authorities through ex ante fiscal rules that provide numerical targets or ceilings for budgetary aggregates or criteria for the conduct of fiscal policy (Krogstrup and Wyplosz (2010)).” Moreover, Tapsoba (2012) uses impact evaluation methods to assess the treatment effect of fiscal rules on budgetary discipline in 74 developing countries over the period 1990-2007 and suggests that fiscal rules are a credible remedy against budgetary indiscipline in developing countries. More recently, Guerguil et al. (2017) assess the impact of different types of flexible fiscal rules on the pro-cyclicality of fiscal policy using propensity scores-matching techniques. The authors find that not all fiscal rules have the same impact: the design matters. Specifically, they find that investment-friendly rules reduce the pro-cyclicality of both overall and investment spending. The effect appears stronger in bad times and when the rule is enacted at the national level. In addition to fiscal rules, recent papers have shed light on the role of institutional quality in pro- moting fiscal discipline: good institutions have been proposed as a remedy against fiscal policy pro-cyclicality (Frankel et al. (2013); Bergman and Hutchison (2015)). For example, Muscatelli et al. (2012) explain that transparency in the decision-making process helps achieve fiscal solvency. on et al. (2012) examine the role of the level of institutional quality on governments ability Calder´ to deliver counter-cyclical macroeconomic policies. Using a global sample of 115 industrial and developing countries over the time span 1984-2008, they find that institutional quality plays a key role in countries’ capacity to implement counter-cyclical macroeconomic policies. Their results show that countries with strong (weak) institutions adopt counter- (pro) cyclical macroeconomic policies. Frankel et al. (2013) show that good institutions fundamentally explain the reduction of policy pro-cyclicality. They find that countries graduate from pro-cyclicality as institutions im- prove. Moreover, Foremny (2014) shows that the effectiveness of fiscal rules and tax autonomy depends on the constitutional structure of government. The author further explains that fiscal ag- gregates vary greatly from country to country, even for countries with similar economic conditions, 7 and differences in deficits can be attributed to differences between countries in terms of political and institutional factors. Iara and Wolff (2014) show that the legal base of fiscal rules and their enforcement mechanisms are the most important dimensions of rule-based fiscal governance. Fur- thermore, Bergman and Hutchison (2015) underline, in a dynamic panel model with 81 developed, emerging and developing countries, that fiscal rules alone are not enough to reduce pro-cyclicality and should be combined with better institutions in order to become more effective. Our paper attempts to test 3 hypotheses concerning fiscal rules, institutional quality, exchange rate regimes and the cyclicality of fiscal policy. H1: Better institutional quality promotes counter-cyclical fiscal policy, thus having a beneficial effect on fiscal discipline. H2: Fiscal rules contribute to promote counter-cyclical fiscal policy and are more effective when implemented in an environment with strong institutions. Rules and good institutions are comple- mentary. H3: The effect of fiscal rules on the cyclicality of fiscal policy differs across alternative types of exchange rate regimes. 3 Methodology As aforementioned, our research question is the following: to what extent do fiscal rules, institutions and exchange rate regimes affect the cyclical behavior of fiscal policy? We implement a two- step approach using a specification which first deals with the cyclicality of fiscal policy and then examines the effects of rules, institutions and exchange rate regimes and their interactive effects on cyclicality. In the first step, we follow Aghion et al. (2007) and Guerguil et al. (2017) by computing cyclicality coefficients of fiscal policy using a non-parametric regression method: the Local Gaussian-Weighted 8 Ordinary Least Squares (LGWOLS). This approach allows computing cyclicality coefficients that are country-specific and time-varying. Indeed, it allows capturing the fact that government reaction to business cycle fluctuations may vary over time and differ between the up and down phases of the business cycle. The estimation of equation (1) uses LGWOLS and allows computing the cyclicality coefficients. ∆Log (GEXP it ) = αit + βit ∆Log (GDP it ) + εit (1) 2 exp (− (τ2− σ 2 1 t) With εit → N (0, ωt (τ ) ) and ωt (τ ) = √ σ 2π σ2 ) ˆit , the LGWOLS weights all observations by a Gaussian centered at t, To compute coefficients β for country i and then performs one regression for each date t. In fact, the method uses all the observations for each year and the closest observations to the year considered are given a greater weight. Concerning the choice of σ , we follow Aghion et al. (2007) and Guerguil et al. (2017) and use a value of the parameter σ equal to 5. ∆Log (GEXP it ) is the growth rate of government real expenditure for country i at time t and ∆Log (GDP it ) refers to the growth rate of government real GDP for country i at time t. The ˆit captures the cyclical behavior of public predicted country-specific and time-varying coefficient β spending. It measures the cyclicality of public spending for country i at time t. Hence, fiscal policy ˆit > 0 (resp.< 0) and acyclical otherwise. is considered pro-cyclical (resp. counter-cyclical) if β ˆit (the cyclicality of fiscal policy) as dependent In the second step, we use the predicted coefficients β variable and search for the effects of our interest variables, in relation to the three hypotheses (H1, H2, H3) mentioned in the previous section. Equation (2) puts forward the independent effects of fiscal rules, institutions and exchange rate regimes on cyclicality. It is meant to test H1. Equation (3) allows us capturing the interaction effect of fiscal rules and institutions on cyclicality. Hence, it can capture the insights of H2. Equation (4) in turn allows measuring the effect of fiscal rules conditional on the exchange rate regime in place: it tackles thus H3. 9 ˆit = αit + λF RDit + δBQ t + θERRit + β φk Xkit + ηi + εit (2) i ˆit = αit + λF RDit + δBQ t + θERRit + ϕF RDit × BQ t + β φk Xkit + ηi + εit (3) i i ˆit = αit + λF RDit + δBQ t + θERRit + ρF RDit × ERRit + β φk Xkit + ηi + εit (4) i Overall, equations (1), (2) and (3) allow testing respectively the three hypotheses (H1, H2 and H3) mentioned in the literature section. F RDit is the fiscal rules dummy for country i at time t and BQi t is our institutional variable, the bureaucracy quality index for country i at time t. ERRit is the variable of exchange rate regimes classification. In our analysis, we first run the equations (2) (3) and (4) for the presence of fiscal rules in general. Then we consider the presence of different types of fiscal rules since numerical rules could have different budgetary outcomes according to the type of budgetary aggregate which is concerned. We differentiate between expenditure rules (ER), debt rules (DR), budget balance rules (BBR) and revenue rules (RR). The coefficient λ stands for the direct effect of fiscal rules on expenditure cyclicality. δ measures the effect of institutions on cyclicality and the coefficient θ measures the effect of exchange rate regimes on cyclicality. The coefficient ϕ captures the interactive effect of fiscal rules and institutions. In turn, ρ allows measuring the effect of fiscal rules conditioned to the exchange rate regime in place. αit the constant term is the cyclicality of expenditures which is independent from any of the explanatory variables. ηi stands for the country fixed effects and εit is the idiosyncratic error. We also add a vector of covariates X in order to control for other variables that could affect the conduct of fiscal policy according to the literature. The set of control variables includes an inflation targeting dummy, the number of years prior to elections, an index for government polarization, the output volatility, the logarithm of GDP per capita, the ratio of public debt to GDP and the ratio of natural resource rents to GDP. We perform the estimation of our second-step equations (namely equations (2), (3) and (4)) using first, basic panel regression (estimation results in section 5). And then in section 6, we perform the regressions using an instrumental variable technique (2SLS) to address endogeneity concern. 10 4 Data and Statistical Outline Our investigation covers a total of 153 countries over a 23-years period: 1993-2015. To measure the cyclicality of fiscal policy, we compute, as aforementioned, the impact of real GDP growth on real expenditure growth using the LGWOLS. Our preference for public expenditure as fiscal policy variable rather than others like primary balance or revenue is motivated by the criticism of Kaminsky et al. (2004). The authors argue that fiscal balance or revenue are results of fiscal policy rather than instruments. In fact, if a government wants to influence economic activity, it would modify its expenditure program or change tax rates and such effects would affect revenue and fiscal balance. Moreover, the advantage of using government expenditures is that co-movements with GDP are able to clearly differentiate between pro-cyclical, a-cyclical and counter-cyclical fiscal policy (Bergman and Hutchison (2015)). The source of the data on fiscal rules is “IMF Fiscal Rules Dataset 2016” from the Fiscal Affairs Department of the International Monetary Fund. This dataset contains information on all the types of fiscal rules (expenditure rules, debt rules, budget balance rules, revenue rules) and their characteristics for the countries in which they did exist at any time since 1985. We perform our regressions using dummy variables of different types of fiscal rules. FRD stands for fiscal rules dummy, ERD for expenditure rules dummy, DRD is debt rules dummy, BBRD is budget balance rules dummy and RRD is revenue rules dummy. Our institutional quality variable is “Bureaucracy quality” from International Country Risk Guide (ICRG). Bureaucracy quality represents “the institutional strength and quality of the bureaucracy; a shock absorber that tends to minimize revisions of policy when governments change” (ICRG 2015). It ranges between 0 and 4, with “high points indicating countries where the bureaucracy has the strength and expertise to govern without drastic changes in policy or interruptions in government services. In these low-risk countries, the bureaucracy tends to be somewhat autonomous from po- litical pressure and to have an established mechanism for recruitment and training” (ICRG). 11 We use as exchange rate regimes variable, the classification by Ilzetzki et al. (2017). The classifi- cation ranges from 1 to 15, with lower values representing more fixed exchange rate regimes and higher values representing more flexible regimes (See Appendix 4). As aforementioned, the set of control variables that we employ includes an inflation targeting dummy, the number of years prior to elections, an index for government polarization, the output volatility, the logarithm of GDP per capita, the ratio of public debt to GDP and the ratio of natural resource rents to GDP. The use of inflation targeting is motivated by the fact that this institutional reform may influence the conduct of fiscal policy by limiting the scope of seigniorage revenues and placing other constraints on expenditure (Combes et al. (2017)). To address this point, we use a dummy variable equal to 1 if the country has implemented inflation targeting at time t and 0 otherwise. The data are taken from Aizenman et al. (2011), Samarina and De Haan (2014), and Laurens et al. (2015). In addition, we take account of the electoral cycles to explain fiscal discipline since the main theoretical explanations of fiscal indiscipline emerge from the political budget cycle models. We use the variable elections which is the number of years prior to executive elections. We construct this variable using the database of Hyde and Marinov (2012): The National Elections across Democracy and Autocracy (NELDA) data set. The literature on government fragmentation suggests that fragmentation has harmful effect on pub- lic finances (Kontopoulos and Perotti (1999); Volkerink and De Haan (2001)). Thus, government polarization may be associated with higher fiscal discipline. We use the index Polariz from the World Bank Database of Political Institutions to control for government polarization. It represents the maximum polarization between the executive party and the four principle parties of the legis- lature. The higher the index, the more polarized the government is. The index is zero if the chief executive’s party has an absolute majority in the legislature. Moreover, we follow Frankel et al. (2013) and Talvi and Vegh (2005) by controlling for the variability of tax revenues that we proxy using output variability. This channel has been emphasized by 12 Talvi and Vegh (2005) who argue that, in the presence of political distortions, the larger the variability of tax revenues, the more pro-cyclical fiscal policy will be. We use the square of the cyclical component of real GDP from Hp filter (parameter 6.5) to measure output variability. We also control for the characteristics inherent to the level of development of countries using the logarithm of GDP per capita. This allows to know whether the behavior of fiscal policy depends on the level of development. We control for the level of public debt. Controlling for this variable seems somewhat compul- sory since public debt is an important element of decision when designing public expenditure programs. Finally, the literature on Dutch disease suggests that the abundance of natural resources in a particular country and the mismanagement of resource rents could have detrimental effects on public finances (Torvik (2009)). Hence, we use the ratio of natural resource rents to GDP as a control variable. Appendix 1 recapitulate the variables used and the source of data. Appendix 2 displays basic descriptive statistics on the whole set of variables employed in the model. We also add at the bottom of the table, descriptive statistics on literacy rate of adults (LRA) and life expectancy at birth (LEB), (obtained from World Bank WDI), that we use, in the section 6, as instruments for institutional quality, in addition to the lagged variables of institutional quality. The correlation matrix in Appendix 3 reports the degree of correlation between the different vari- ables. It suggests that multicolinearity is not likely to be an issue and hence authorizes the intro- duction of the different variables in the empirical model. Since our variable of cyclicality of fiscal policy is generated using LGWOLS, we display in figure 1 the Kernel density plot to view the distribution of the variable. On the right, the figure presents the same plot for the distribution of bootstrapped estimates of the cyclicality (with 200 replications). The distribution of the generated cyclicality coefficients is roughly normal. 13 Figure 1: Density plot of cyclicality coefficients 5 Empirical Results Table 1 reports the estimation results using basic panel regression for equations 2, 3 and 4. We use as fiscal rules variable the dummy variable reflecting the existence of any type of fiscal rules. We search for the effect of exchange rate regimes, the adoption of fiscal rules, and the institutional quality on the cyclical behavior of fiscal policy. In column 9, we add all the control variables and the interactive variable of fiscal rules and institutional quality. In column 10, we add the interactive effect of fiscal rules and exchange rate regimes in order to examine the effect of fiscal rules conditioning it on the type of exchange rate regime in place. We do not find a significant beneficial effect of the adoption of fiscal rules on the cyclical behavior of fiscal policy. However, in line with Frankel et al. (2013) and Bergman and Hutchison (2015), institutional quality has beneficial effects on the conduct of fiscal policy. The better is the in- stitutional quality, the more counter-cyclical (or less pro-cyclical) fiscal policy is. The coefficient of the interactive term between fiscal rules and institutional quality is negative and statistically significant. This suggests that when institutions are strong, fiscal rules are effective in promoting counter-cyclical fiscal policy. Good institutions combined with fiscal rules have a disciplining effect on fiscal policy. The more institutions improve, the more effective are fiscal rules in promoting 14 counter-cyclical policies, and thus fiscal discipline. Numerical fiscal rules and good institutions form a potent combination that has a beneficial effect on fiscal discipline. This finding is line with Bergman and Hutchison (2015) and Wyplosz (2012) which stipulate that to be effective, rules must be embedded in institutional arrangements. Moreover, we find that the more fixed (resp. flexible) the exchange rate regime is, the more counter- cyclical (resp. pro-cyclical) fiscal policy is. Hence, more pegged exchange rate regimes promote counter-cyclical fiscal policy, having thus a disciplining fiscal effect. This finding is in line with the conventional wisdom (Canavan and Tommasi (1997) Ghosh et al. (2010) and Sow (2015)). When it comes to analyzing the interaction effect of fiscal rules with the exchange rate regime, we find a negative coefficient, stipulating that under more flexible exchange rate regimes, fiscal rules have significant beneficial effect on the conduct of fiscal policy. The more flexible an exchange rate regime is, the more effective fiscal rules are in promoting counter-cyclical policies. However, under fixed regimes, rules seem to become counter-productive. This can be explained by the fact that pegged exchange rate regimes constrain the use of monetary policy tools and let governments have hands only on the fiscal policy tools, namely on public spending. Hence, by constraining fiscal policy, the adoption of fiscal rules results in the doubling of the restrictions affecting governments and this could not yield disciplined behavior. It would instead yield more discretionary and corrupted behavior since governments are doubly constrained (at the monetary policy level and at the fiscal policy level). Our empirical results also show that inflation targeting is beneficial for fiscal discipline. In fact, this institutional reform may influence the conduct of fiscal policy by limiting the scope of seignior- age revenues and placing other constraints on expenditure. Moreover, we find that government polarization has beneficial effect on fiscal discipline. When fiscal policy is decided on a centralized basis, the voracity effects due to the conflicts of interest of many groups could disappear and in- stead yield inter-temporal smoothing behaviors which are better for macroeconomic stability. The findings also suggest that the electoral cycle significantly affects the conduct of fiscal policy. The higher the number of years prior to executive elections is, the more pro-cyclical fiscal policy is. 15 When elections are far, fiscal policy is less disciplined and when countries are close to executive elections, governments tend to groom electors by performing more disciplined policies in order to guarantee their reelection. GDP per capita is negatively associated with cyclicality coefficients. This means that more developed countries tend to perform counter-cyclical fiscal policies. Further- more, we find that the volatility of output yields pro-cyclical policy as emphasized by Talvi and Vegh (2005). Dependent variable: Cyclicality of fiscal policy βˆ it (1) (2) (3) (4) (5) (6) (7) (8) (9) (10) FRD -0.058*** -0.016*** 0.060 0.144 0.197 0.197 0.180 0.069 0.300 0.083 (0.032) (0.033) (0.250) (0.244) (0.241) (0.239) (0.239) (0.249) (0.593) (0.490) BQ -0.194*** -0.183*** -0.091*** -0.111*** -0.032*** -0.034*** -0.132*** -0.021*** 0.020*** -0.020*** (0.033) (0.033) (0.050) (0.059) (0.061) (0.060) (0.063) (0.073) (0.070) (0.071) ERR 0.005*** 0.004*** 0.009*** 0.012*** 0.007*** 0.007*** 0.006*** 0.008*** 0.008*** 0.009*** (0.004) (0.004) (0.050) (0.012) (0.011) (0.011) (0.011) (0.010) (0.010) (0.014) FRD X BQ -0.107*** (0.070) FRD X ERR -0.002*** (0.015) IT -0.253*** -0.441*** -0.618*** -0.586*** -0.585*** -0.528*** -0.447*** -0.456*** -0.447*** (0.051) (0.086) (0.106) (0.107) (0.106) (0.129) (0.053) (0.052) (0.055) ELECTIONS 0.003*** 0.002*** 0.001*** 0.001*** 0.000*** 0.004*** 0.004*** 0.004*** (0.003) (0.003) (0.003) (0.003) (0.003) (0.003) (0.003) (0.003) POLARIZ -0.149*** -0.148*** -0.147** * -0.177*** -0.099*** -0.101*** -0.099*** (0.014) (0.015) (0.015) (0.016) (0.010) (0.010) (0.010) GDPPC -0.274*** -0.276*** -0.236*** -0.193*** -0.172*** -0.191*** (0.025) (0.024) (0.022) (0.024) (0.030) (0.024) OV 5.606** 15.211*** 19.987 20.875 19.953 (3.763) (3.414) (6.854) (7.231) (7.183) RENT -0.023*** -0.029** -0.028*** -0.029*** (0.003) (0.003) (0.003) (0.003) DEBT 0.001 0.001 0.001 (0.001) (0.001) (0.001) Obs 2264 2264 1288 1021 1004 1004 985 912 912 912 Countries 120.000 120.000 72.000 71.000 70.000 70.000 69.000 68.000 68.000 68.000 Hansen J 0.696 0.698 0.662 0.711 0.703 0.674 0.690 0.732 0.735 0.736 Bootstrapped standard errors in parenthesis (100 replications) * p<0.10, ** p<0.05, *** p<0.01 Table 1: Empirical results: basic panel regression 16 In figure 2, we plot the cyclicality coefficients against the values of the interactive term of fiscal rules and institutions.3 . The illustrative figure confirms that rules and good institutions together have beneficial effects on fiscal discipline since they promote counter-cyclical macroeconomic poli- cies. Figure 2: Cross-correlation of fiscal policy cyclicality with the interactive term of FRD and BQ 3 Figure 2 is a binned scatterplot. We compute it using the stata command binscatter. Binned scatterplots provide a non-parametric way of visualizing the relationship between two variables. With a large number of observations, a scatterplot that plots every data point would become too crowded to interpret visually. The Stata command binscatter groups the x-axis variable into equal-sized bins, computes the mean of the x-axis and y-axis variables within each bin, then creates a scatterplot of these data points. The result is a non-parametric visualization of the conditional expectation function. It is optimized for speed in large data sets. 17 In order to refine our investigation, we specifically examine how different types of rules influence the cyclicality of fiscal policy. We differentiate between expenditure rules (ER), budget balance rules (BBR), debt rules (DR) and revenue rules (RR). Table 2 reports the estimation results for expenditure rules, budget balance rules and debt rules. Specifically, we find that while expenditure rules and budget balance rules seem to be associated with fiscal pro-cyclicality, debt rules and revenue rules seem to favor more budgetary discipline. Dependent variable: Cyclicality of fiscal policy βˆ it (1) (2) (3) (4) BQ -0.028*** -0.029*** -0.058*** -0.041*** (0.087) (0.071) (0.062) (0.088) ERR 0.008*** 0.008*** 0.007*** 0.008*** (0.011) (0.010) (0.011) (0.010) ERD 0.235** (0.066) BBRD 0.021*** (0.081) DRD -0.155*** (0.102) RRD -0.130** (0.117) IT -0.427*** -0.437*** -0.420*** -0.435*** (0.054) (0.054) (0.055) (0.053) ELECTIONS 0.003*** 0.004*** 0.004*** 0.005*** (0.003) (0.003) (0.003) (0.003) POLARIZ -0.096*** -0.099*** -0.097*** -0.100*** (0.010) (0.009) (0.010) (0.010) GDPPC -0.204*** -0.183*** -0.147*** -0.174*** (0.033) (0.023) (0.032) (0.032) OV 19.208 19.636 17.990 18.513 (6.865) (6.915) (5.841) (6.149) RENT -0.029*** -0.029*** -0.029*** -0.029*** (0.003) (0.003) (0.003) (0.003) DEBT 0.001 0.001 0.001 0.001 (0.001) (0.001) (0.001) (0.001) Obs 912 912 912 912 Countries 68.000 68.000 68.000 68.000 Hansen J 0.736 0.728 0.733 0.735 Bootstrapped standard errors in parenthesis (100 replications) * p<0.10, ** p<0.05, *** p<0.01 Table 2: Results for ER, BBR, DR, RR 18 6 Addressing Endogeneity In this section, we perform the estimations of our second-step equations, (namely equations (2), (3) and (4)) using an instrumental variable technique: the Two-Stage Least Squares (2SLS) esti- mator in order to address potential endogeneity problem. Table 3 provides the estimation results that count for endogeneity. In fact, one could argue that counter-cyclical (resp. pro-cyclical) fis- cal policies that tend to stabilize (resp. destabilize) the economy might improve (resp. worsen) institutional quality (Frankel et al. (2013)). “For example, procyclical fiscal policies could increase the chances of governments running into debt sustainability problems during busts. These critical financing needs could then lead to expropriation, repudiation of contracts, and/or intervention in independent branches of governments such as the judiciary system or the central bank. Moreover, the turmoil typically associated with debt crises can exacerbate corruption in the political system thus weakening the foundations of an efficient and professional public administration.” 4 That is to say, the causality may run from cyclicality of fiscal policies to institutional quality and the re- verse way around. To address such endogeneity concern, we instrument institutional quality. The literature on institutions has not yet found time-varying instrumental variables for the quality of institutions and relies on cross-country regressions and instruments institutions using European settlers’ mortality and latitude (absolute value): Acemoglu et al. (2001), Frankel et al. (2013). Since we are in panel, we use as instruments the life expectancy at birth, the literacy rate of adults and the lagged levels of institutional quality. Life expectancy at birth and literacy rate undoubtedly reflect the same information as the ones used by Acemoglu et al. (2001), but have a time varying dimension. 4 Frankel et al. (2013). 19 Dependent variable: Cyclicality of fiscal policy βˆit (1) (2) (3) (4) (5) (6) (7) (8) (9) (10) FRD 0.294 0.326 0.385 0.386 0.555 0.555 0.750 0.733 2.010 2.092 (0.288) (0.017) (0.025) (0.057) (0.388) (0.394) (0.418) (0.423) (1.570) (0.940) BQ -0.602*** -0.589*** -0.578** -0.621* -0.579** -0.578** -0.503 -0.521 * -0.079 * -1.002* (0.310) (0.034) (0.065) (0.208) (0.075) (0.075) (0.117) (0.090) (0.094) (0.090) ERR 0.040*** 0.043*** 0.047** 0.069* 0.075** 0.075** 0.103** 0.101 * 0.118 * 0.252 * (0.039) (0.002) (0.006) (0.022) (0.008) (0.008) (0.008) (0.009) (0.009) (0.010) FRD X BQ -0.549 ** (0.038) FRD X ERR -0.241*** (0.007) IT -0.269 -0.398*** -0.474*** -0.336*** -0.336*** -0.369*** -0.379*** -0.475*** -0.532 *** (0.019) (0.032) (0.106) (0.046) (0.046) (0.025) (0.026) (0.026) (0.026) ELECTIONS 0.033*** 0.030 0.026*** 0.026*** 0.026*** 0.027 *** 0.023 *** 0.031 *** (0.002) (0.004) (0.003) (0.002) (0.003) (0.002) (0.002) (0.002) POLARIZ -0.409*** -0.437*** -0.436*** -0.350*** -0.333*** -0.299*** -0.351*** (0.018) (0.011) (0.011) (0.009) (0.007) (0.007) (0.007) GDPPC -0.892*** -0.890*** -0.808*** -0.729 *** -0.790 *** -0.847*** (0.057) (0.057) (0.055) (0.058) (0.058) (0.059) OV 1.585** 1.833* 13.458 ** 10.062 ** -2.864 ** (3.700) (5.331) (4.481) (4.481) (4.569) DEBT -0.001 -0.002 -0.001 -0.002 (0.001) (0.000) (0.000) (0.000) RENT -0.012 *** -0.008 *** -0.015*** (0.002) (0.002) (0.002) Obs. 2,177 2,177 1,242 978 962 970 901 883 883 883 Countries 120 120 72 71 70 70 61 68 68 68 Hansen J 0.275 0.300 0.240 0.132 0.163 0.163 0.139 0.167 0.153 0.265 Bootstrapped standard errors in parenthesis (100 replications) * p<0.10, ** p<0.05, *** p<0.01 Table 3: Estimation results: addressing endogeneity 7 Robustness Check In this section, we provide a robustness check for our empirical findings. In order to check the consistency of our empirical results, we use an alternative variable of institutional quality from a source other than ICRG. We employ the government effectiveness index from the Worldwide Governance Indicators, 2015 Update database of the World Bank.5 The Government effectiveness reflects “perceptions of the quality of public service, the quality of civil service and the degree of independence of political pressures, the quality of policy formulation and implementation, and the credibility of the government’s commitment to these services.” It was initially comprised between -2.5 and 2.5 with high values indicating more efficient governments. However, we normalize it between 0 and 5. Table 4 displays the estimation results when we perform the same regressions as in the previous 5 The whole data set comprises data on the quality of governance provided by “a large number of enterprise, citizen and expert survey respondents in industrial and developing countries. These data are gathered from a number of survey institutes, think tanks, non-governmental organizations, international organisations, and private sector firms.” 20 section while substituting bureaucracy quality by government effectiveness. The use of that other variable of institutional quality does not fundamentally change our findings. Specifically, we find that fiscal rules may help promote counter-cyclical fiscal policy but are more effective when imple- mented in an environment with good institutions. Fiscal rules, together with strong institutions form a potent combination which promotes counter-cyclical policy. Moreover, we get the same results with the control variables. Inflation targeting and government polarization are associated with more fiscal discipline. Less developed countries tend to perform pro-cyclical policies; and in the hedge of elections, governments tend to perform budgetary disci- pline. Dependent variable: Cyclicality of fiscal policy βˆit (1) (2) (3) (4) (5) (6) (7) (8) (9) 10 FRD -0.022 *** -0.002*** -0.001*** 0.174 0.186 0.188 0.196 0.186 0.730 0.131 (0.007) (0.008) (0.013) (0.011) (0.010) (0.010) (0.010) (0.096) (0.546) (0.086) GE -0.039 *** -0.013*** 0.246*** -0.232** -0.148*** -0.147** -0.176*** -0.122 *** -0.049 * -0.135 *** (0.028) (0.029) (0.070) (0.095) (0.094) (0.097) (0.099) (0.132) (0.200) (0.132) ERR 0.020** 0.022*** 0.036* 0.022* 0.023* 0.023* 0.022* 0.017 *** 0.019 *** 0.014*** (0.004) (0.004) (0.008) (0.013) (0.013) (0.013) (0.013) (0.011) (0.012) (0.011) FRD X GE -0.205* (0.130) FRD X ERR 0.010 (0.013) IT -0.238** -0.244* -0.387** -0.342*** -0.341*** -0.345*** -0.360*** -0.346*** -0.352*** (0.042) (0.068) (0.042) (0.041) (0.041) (0.042) (0.060) (0.062) (0.061) ELECTIONS 0.011*** 0.002*** 0.003*** 0.003*** 0.003 *** 0.002 *** 0.003 *** (0.002) (0.002) (0.002) (0.002) 0.002*** (0.001) (0.001) (0.001) POLARIZ -0.019** -0.029*** -0.031*** (0.002) -0.012*** -0.009 *** -0.010 *** (0.017) (0.016) (0.016) (0.017) (0.014) (0.017) GDPPC -0.303*** -0.308*** -0.306*** -0.268*** -0.273*** -0.267*** (0.051) (0.054) (0.061) (0.090) (0.106) (0.088) OV -5.772* -6.542* -4.509 -6.162 -4.145 (2.675) (2.655) (3.574) (3.338) (5.424) DEBT -0.001*** -0.001*** -0.001 -0.001 *** (0.000) (0.001) (0.001) (0.001) RENT -0.020 -0.018** -0.020*** (0.008) (0.008) (0.007) Obs. 1,316 1,316 724 493 487 487 486 479 479 479 Countries 137 137 78 70 69 69 69 68 68 68 Hansen J 0.269 0.288 0.204 0.816 0.962 0.966 0.982 0.964 0.949 0.963 Bootstrapped standard errors in parenthesis (100 replications) * p<0.10, ** p<0.05, *** p<0.01 Table 4: Robustness check 21 8 Conclusion and Policy Implications We explore how numerical fiscal rules, exchange rate regimes, and institutional quality affect the cyclical behavior of fiscal policy (i.e, how government spending responds to GDP fluctuations). We perform our analysis using a panel of 153 advanced, emerging and developing countries over a 23- year period, from 1993 to 2015. We proceed by a two-step econometric approach. We first compute the cyclicality coefficients of fiscal policy using the Local Gaussian-Weighted Ordinary Least Squares (LGWOLS) and secondly estimate how institutional quality, fiscal rules and exchange rate regimes affect the predicted cyclicality. In addition to basic panel regression, we use an instrumental variable method, the two-stage Least Squares (2SLS), in order to address the endogeneity issue. We find that fiscal rules alone are not enough to promote counter-cyclical fiscal policy and should be combined with better institutions. Fiscal rules, together with strong institutions form a more optimal combination which has a disciplining effect on fiscal policy. This finding is in line with Bergman and Hutchison (2015) and Wyplosz (2012) which stipulate that to be effective, rules must be embedded in institutional arrangements. Our result is robust to the use of alternative variable of institutional quality. Moreover, we show that the more pegged the exchange rate regime is (resp. flexible), the more counter-cyclical (resp. pro-cyclical) fiscal policy is. Hence, more fixed exchange rate regimes pro- mote counter-cyclical fiscal policy, having thus disciplining effects on fiscal policy. Furthermore, we find that under more flexible exchange rate regimes, fiscal rules are effective in promoting counter- cyclical policies. But under more fixed ones, fiscal rules seem to be associated with less fiscal discipline. We also find that the disciplining effect of fiscal rules depends on the type of rule. Specifically, while expenditure rules and budget balance rules seem to be associated with fiscal pro-cyclicality, debt rules and revenue rules seem to favor more budgetary discipline. To sum up, fiscal rules are not a sufficient remedy against fiscal indiscipline and should be com- 22 bined with better institutions. Governments should be encouraged to improve their governance and institutional framework, which is a pledge of budgetary discipline, necessary to accomplish macroe- conomic stabilization. In addition, the prescription or recommendation of fiscal rules seems to be more convenient for countries facing flexibility of their exchange rate regimes than for countries that have pegged exchange rate regimes. 23 References Acemoglu, D., S. Johnson, and J. A. Robinson (2001). The colonial origins of comparative devel- opment: An empirical investigation. American economic review 91 (5), 1369–1401. Aghion, P., I. Marinescu, R. J. Caballero, and A. K. Kashyap (2007). Cyclical budgetary policy and economic growth: What do we learn from oecd panel data?[with comments and discussion]. NBER Macroeconomics annual 22, 251–297. Aizenman, J., M. Hutchison, and I. Noy (2011). Inflation targeting and real exchange rates in emerging markets. World Development 39 (5), 712–724. Alberola, E. and L. Molina (2004). What does really discipline fiscal policy in emerging markets? the role and dynamics of exchange rate regimes. Alesina, A. and G. Tabellini (1990). A positive theory of fiscal deficits and government debt. The Review of Economic Studies 57 (3), 403–414. Alesina, A. and G. Tabellini (2005). Why is fiscal policy often procyclical?, nber working paper, no 11600. Beetsma, R. M. and A. L. Bovenberg (1998). Monetary union without fiscal coordination may discipline policymakers. Journal of international economics 45 (2), 239–258. Bergman, U. M. and M. Hutchison (2015). Economic stabilization in the post-crisis world: Are fiscal rules the answer? Journal of International Money and Finance 52, 82–101. Bergman, U. M., M. M. Hutchison, and S. E. H. Jensen (2016). Promoting sustainable public finances in the european union: The role of fiscal rules and government efficiency. European Journal of Political Economy 44, 1–19. Bova, E., N. Carcenac, and M. Guerguil (2014). Fiscal rules and the procyclicality of fiscal policy in the developing world. 24 Budina, M. N., M. T. Kinda, M. A. Schaechter, and A. Weber (2012). Fiscal rules at a glance: Country details from a new dataset. Number 12-273. International Monetary Fund. on, C., R. Duncan, K. Schmidt-Hebbel, et al. (2012). Do good institutions promote counter- Calder´ cyclical macroeconomic policies? Federal Reserve Bank of Dallas, Globalization and Monetary Policy Institute Working Paper 118. Canavan, C. and M. Tommasi (1997). On the credibility of alternative exchange rate regimes. Journal of Development Economics 54 (1), 101–122. Canzoneri, M. B., R. E. Cumby, and B. T. Diba (2001). Fiscal discipline and exchange rate systems. The Economic Journal 111 (474), 667–690. Christiano, L., M. Eichenbaum, and S. Rebelo (2011). When is the government spending multiplier large? Journal of Political Economy 119 (1), 78–121. Colciago, A., T. Ropele, V. A. Muscatelli, and P. Tirelli (2008). The role of fiscal policy in a monetary union: are national automatic stabilizers effective? Review of International Eco- nomics 16 (3), 591–610. Combes, J.-L., X. Debrun, A. Minea, and R. Tapsoba (2017). Inflation targeting, fiscal rules, and the policy mix: Cross-effects and interactions. The Economic Journal . Corbacho, A. and G. Schwartz (2007). Fiscal responsibility laws. Promoting fiscal discipline 58, 71. Debrun, X., L. Moulin, A. Turrini, J. Ayuso-i Casals, and M. S. Kumar (2008). Tied to the mast? national fiscal rules in the european union. Economic Policy 23 (54), 298–362. Diallo, O. (2009). Tortuous road toward countercyclical fiscal policy: Lessons from democratized sub-saharan africa. Journal of Policy Modeling 31 (1), 36–50. 25 Duttagupta, R. and T. Guillermo (2006). Fiscal discipline and exchange rate regimes: evidence from the caribbean. as, A. and A. K. Rose (2001). Do monetary handcuffs restrain leviathan? fiscal policy in extreme Fat´ exchange rate regimes. Foremny, D. (2014). Sub-national deficits in european countries: The impact of fiscal rules and tax autonomy. European Journal of Political Economy 34, 86–110. Frankel, J. A., C. A. Vegh, and G. Vuletin (2013). On graduation from fiscal procyclicality. Journal of Development Economics 100 (1), 32–47. Gavin, M. and R. Perotti (1997). Fiscal policy in latin america. NBER macroeconomics annual 12, 11–61. Ghosh, M. A. R., M. J. D. Ostry, and M. C. G. Tsangarides (2010). Exchange rate regimes and the stability of the international monetary system. Number 270. International Monetary Fund. Guerguil, M., P. Mandon, and R. Tapsoba (2017). Flexible fiscal rules and countercyclical fiscal policy. Journal of Macroeconomics 52, 189–220. Hyde, S. D. and N. Marinov (2012). Which elections can be lost? Political Analysis 20 (2), 191–210. Iara, A. and G. B. Wolff (2014). Rules and risk in the euro area. European Journal of Political Economy 34, 222–236. Ilzetzki, E., C. M. Reinhart, and K. S. Rogoff (2017). Exchange arrangements entering the 21st century: Which anchor will hold? Technical report, National Bureau of Economic Research. egh (2004). When it rains, it pours: procyclical Kaminsky, G. L., C. M. Reinhart, and C. A. V´ capital flows and macroeconomic policies. NBER macroeconomics annual 19, 11–53. Kontopoulos, Y. and R. Perotti (1999). Government fragmentation and fiscal policy outcomes: 26 Evidence from oecd countries. In Fiscal institutions and fiscal performance, pp. 81–102. University of Chicago Press. Krogstrup, S. and C. Wyplosz (2010). A common pool theory of supranational deficit ceilings. European Economic Review 54 (2), 269–278. e (2015). The Laurens, M. B., K. Eckhold, D. King, M. N. Ø. Mæhle, A. Naseer, and A. Durr´ journey to inflation targeting: Easier said than done the case for transitional arrangements along the road. Number 15-136. International Monetary Fund. Muscatelli, V. A., P. Natale, and P. Tirelli (2012). A simple and flexible alternative to stability and growth pact deficit ceilings. is it at hand? European Journal of Political Economy 28 (1), 14–26. Nakata, H. (2011). Informative advertising with spillover effects. International Journal of Economic Theory 7 (4), 373–386. Persson, T. and L. E. Svensson (1989). Why a stubborn conservative would run a deficit: Policy with time-inconsistent preferences. The Quarterly Journal of Economics 104 (2), 325–345. Samarina, A. and J. De Haan (2014). Right on target: exploring the factors leading to inflation targeting adoption. Contemporary Economic Policy 32 (2), 372–389. Schaechter, A., T. Kinda, N. T. Budina, and A. Weber (2012). Fiscal rules in response to the crisis-toward the’next-generation’rules: A new dataset. Schuknecht, L. (1999). Fiscal policy cycles and the exchange rate regime in developing countries. European Journal of Political Economy 15 (3), 569–580. e Sow, M. N. (2015). Essays on Exchange Rate Regimes and Fiscal Policy. Ph. D. thesis, Universit´ d’Auvergne-Clermont-Ferrand I. Talvi, E. and C. A. Vegh (2005). Tax base variability and procyclical fiscal policy in developing countries. Journal of Development economics 78 (1), 156–190. 27 Tapsoba, R. (2012). Do national numerical fiscal rules really shape fiscal behaviours in developing countries? a treatment effect evaluation. Economic Modelling 29 (4), 1356–1369. Thornton, J. (2008). Explaining procyclical fiscal policy in african countries. Journal of African Economies 17 (3), 451–464. Tornell, A. and P. R. Lane (1999). The voracity effect. american economic review , 22–46. Tornell, A. and A. Velasco (1992). The tragedy of the commons and economic growth: why does capital flow from poor to rich countries? Journal of Political Economy 100 (6), 1208–1231. Tornell, A. and A. Velasco (1995). Fiscal discipline and the choice of exchange rate regime. European Economic Review 39 (3), 759–770. Torvik, R. (2009). Why do some resource-abundant countries succeed while others do not? Oxford Review of Economic Policy 25 (2), 241–256. Volkerink, B. and J. De Haan (2001). Fragmented government effects on fiscal policy: New evidence. Public choice 109 (3), 221–242. Von Hagen, J. and I. J. Harden (1995). Budget processes and commitment to fiscal discipline. European Economic Review 39 (3), 771–779. Wyplosz, C. (2005). Fiscal policy: institutions versus rules. National Institute Economic Re- view 191 (1), 64–78. Wyplosz, C. (2012). Fiscal rules: Theoretical issues and historical experiences. In Fiscal policy after the financial crisis, pp. 495–525. University of Chicago Press. 28 9 Appendix Variable Description Source GEXP Government real expenditure IMF WEO April 2017 GDP Government real GDP IMF WEO April 2017 βˆ it Cyclicality of real expenditure Estimation of equation (1) FRD Fiscal rules dummy (0;1) IMF Fiscal rules dataset 2017 ERD Expenditure rules dummy (0;1) IMF Fiscal rules dataset 2017 BBRD Budget balance rules dummy (0;1) IMF Fiscal rules dataset 2017 DRD Debt rules dummy (0;1) IMF Fiscal rules dataset 2017 RRD Revenue rules dummy (0;1) IMF Fiscal rules dataset 2017 BQ Bureaucracy quality [0;4] ICRG Dataset 2015 ERR Exchange rate regimes [0;15] Ilzetzki et al. (2017) and cf. Appendix 4 IT Inflation targeting Dummy Aizenman et al. (2011), Samarina and De Haan (2014), Laurens et al. (2015) cf. Appendix 5 ELECTIONS Years prior to elections Hyde and Marinov (2012); our calculations POLARIZ Polarization index [0;2] World Bank Database of political institutions 2015 OV Output variability IMF WEO April 2017; our calculations GDPPC Logarithm of GDP per capita World Bank WDI DEBT Ratio of public debt to GDP IMF WEO April 2017 RENT Ratio of natural resource rent to GDP World Bank WDI Table 5: Appendix 1 Description of variables and source 29 Variable Obs Mean Std Dev. Min Max βˆit 3073 0.451 1.257 -4.685 6.454 FRD 3519 0.370 0.483 0 1 BQ 2575 2.267 1.131 0 4 ERR 3519 6.205 4.378 0 15 IT 3519 0.144 0.351 0 1 ELECTIONS 1953 2.412 2.243 0 17 POLARIZ 2470 0.426 0.775 0 2 OV 3476 0.000 0.006 0 0.351 GDPPC 3339 8.431 1.521 4.749 11.61 DEBT 3009 56.95 48.94 0 789.8 RENT 3556 7.580 11.95 0 89.16 LRA 449 80.72 21.22 10.89 99.97 LEB 3449 68.67 9.478 43.17 83.84 Table 6: Appendix 2 Descriptive statistics βˆ it frd bq err it elections polariz ov gdppc debt rent βˆit 1.00 frd -0.05 1.00 bq -0.34 0.39 1.00 err 0.00 -0.14 -0.17 1.00 it -0.13 0.22 0.31 -0.11 1.00 elections -0.04 0.02 0.06 -0.01 -0.06 1.00 polariz -0.01 0.02 0.03 -0.02 -0.02 -0.08 1.00 ov 0.02 -0.03 -0.06 0.01 -0.03 0.01 -0.03 1.00 gdppc -0.32 0.40 0.78 -0.07 0.30 -0.03 0.03 -0.02 1.00 debt 0.13 0.02 -0.06 -0.07 -0.12 0.07 -0.05 0.08 -0.15 1.00 rent -0.01 -0.18 -0.39 0.00 -0.15 0.11 -0.06 0.15 -0.22 0.01 1.00 Table 7: Appendix 3 Correlation Table 30 Classification Bucket of Ilzetzki et al. (2017) Code No separate legal tender 1 Pre announced peg or currency board arrangement 2 Pre announced horizontal band that is narrower than or equal to +/-2% 3 De facto peg 4 Pre announced crawling peg 5 Pre announced crawling band that is narrower than or equal to +/-2% 6 De factor crawling peg 7 De facto crawling band that is narrower than or equal to +/-2% 8 Pre announced crawling band that is wider than or equal to +/-2% 9 De facto crawling band that is narrower than or equal to +/-5% 10 Moving band that is narrower than or equal to +/-2% (i.e., allows for 11 both appreciation and depreciation over time) Managed floating 12 Freely floating 13 Freely falling 14 Dual market in which parallel market data is missing. 15 Table 8: Appendix 4 Classification of exchange rate regimes using Ilzetzki et al. (2017) 31 Country Year of adoption Year of abandonment Albania 2009 . Armenia 2006 . Australia 1993 . Brazil 1999 . Canada 1991 . Chile 1991 . Colombia 1999 . Czech Republic 1998 . Dominican Republic 2012 . Finland 1993 2000 Ghana 2007 . Guatemala 2005 . Hungary 2001 . Iceland 2002 . India 2015 . Indonesia 2005 . Israel 1992 . Korea, Rep. 1998 . Mexico 1999 . Moldova 2010 . New Zealand 1990 . Norway 2001 . Paraguay 2011 . Peru 1994 . Philippines 2002 . Poland 1998 . Romania 2005 . Slovak Republic 1995 2009 Spain 1993 2000 South Africa 2000 . Sweden 1993 . Switzerland 2000 . Turkey 2006 . Uganda 2011 . United Kingdom 1993 . United States 2012 . Table 9: Appendix 5 List of countries that adopted inflation targeting 32 Country name Adoption of rules Country name Adoption of rules Albania no Djibouti no Algeria no Dominica yes Antigua and Barbuda yes Dominican Republic no Argentina yes Ecuador yes Armenia yes Egypt, Arab Rep. no Australia yes El Salvador no Austria yes Equatorial Guinea yes Azerbaijan, Rep. no Estonia yes Bahamas, The no Eswatini no Bahrain, Kingdom of no Ethiopia no Bangladesh no Finland yes Barbados no France yes Belarus no Gabon yes Belgium yes Gambia, The no Belize no Georgia yes Benin yes Germany yes Bhutan no Ghana no Bolivia no Greece yes Botswana yes Grenada yes Brazil yes Guatemala no Brunei Darussalam no Guinea no Bulgaria yes Guinea-Bissau yes Burkina Faso yes Guyana no Burundi yes Haiti no Cameroon yes Honduras no Canada yes Hungary yes Central African Republic yes Iceland yes Chad yes India yes Chile yes Indonesia yes China no Iran, Islamic Rep. yes Colombia yes Ireland yes Congo Rep. yes Israel yes Congo, Dem. Rep. no Italy yes Costa Rica yes Jamaica yes ote d’Ivoire Cˆ yes Japan yes Croatia yes Jordan no Cyprus yes Kazakhstan no Czech Republic yes Kenya yes Denmark yes 33 Kiribati yes Table 10: Appendix 6 List of countries in the sample and adoption of rules Country name Adoption of rules Country Name Adoption of rules Korea, Rep. no Poland yes Kuwait no Portugal yes Kyrgyz Republic no Qatar no Latvia yes Romania yes Lebanon no Russian Federation yes Lesotho no San Marino no Liberia yes Saudi Arabia no Libya no Senegal yes Lithuania yes Singapore yes Luxembourg yes Slovak Republic yes North Macedonia no Slovenia yes Madagascar no South Africa no Malaysia yes Spain yes Maldives yes Sri Lanka yes Mali yes St. Kitts and Nevis yes Malta yes St. Lucia yes Marshall Islands no St. Vincent and the Grenadines yes Mauritania no Sudan no Mauritius yes Suriname no Mexico yes Sweden yes Micronesia, Fed. Sts. no Switzerland yes Moldova no Syrian Arab Republic no Mongolia yes Tajikistan no Morocco no Tanzania yes Myanmar no Togo yes Nepal no Trinidad and Tobago no Netherlands yes Tunisia no New Zealand yes Turkey no Nicaragua no Turkmenistan no Niger yes Uganda yes Nigeria yes Ukraine no Norway yes United Kingdom yes Pakistan yes United States yes Panama yes Uruguay yes Papua New Guinea no Venezuela, RB no Paraguay yes Zambia no Peru yes Zimbabwe no Philippines no 34 Table 11: Appendix 6’ List of countries in the sample and adoption of rules (cont’d)