Policy Research Working Paper 10984 From Fiscal Cyclicality to Fiscal Stress The Role of Asymmetric Public Consumption Rigidities in Emerging Markets Daniel Riera-Crichton Pilar Ruiz Orrico Guillermo Vuletin Latin America and the Caribbean Region Office of the Chief Economist November 2024 Policy Research Working Paper 10984 Abstract Macroeconomic textbooks warn that procyclical public they boost spending during upswings but do not effec- spending can amplify economic volatility and cause fiscal tively cut back during downturns. In contrast, advanced stress. However, the latter risks materialize only when gov- economies maintain steady levels of public consumption ernments fail to reduce spending during downturns as much regardless of economic conditions, making them effectively as they increase it during booms. This study investigates acyclical. Downward rigidity in public consumption not asymmetries in the cyclicality of public consumption and only paves the way for fiscal stress when the economy slows, finds that emerging markets exhibit “downward rigidity”: but also leads to sustained growth in the size of the state. This paper is a product of the Office of the Chief Economist, Latin America and the Caribbean Region. 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 drieracr@bates.edu, pruizorrico@worldbank.org, and gvuletin@worldbank.org. 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 From Fiscal Cyclicality to Fiscal Stress: The Role of Asymmetric Public Consumption Rigidities in Emerging Markets Daniel Riera-Crichton, Pilar Ruiz Orrico, Guillermo Vuletin1 Keywords: Business Cycle; Cyclicality; Downward Rigidity; Public Consumption JEL: E320; E620; H500 1 Riera-Crichton: Bates College; email: drieracr@bates.edu Ruiz Orrico: World Bank; email: pruizorrico@worldbank.org Vuletin: World Bank; email: gvuletin@worldbank.org 1. Introduction Developing markets often grapple with fiscal challenges that can unexpectedly be traced to periods of economic prosperity. In these "good times” buoyed by robust economic growth, public revenues surge, while financing costs remain relatively low due to renewed market confidence. Consequently, emerging markets loosen the constraints on their spending. However, beyond the immediate benefits and drawbacks of these procyclical policies—such as the potential for expanded provision of essential public goods and increased economic volatility (Kaminsky, Reinhart and Végh 2005)—an underlying concern arises. Emerging markets tend to allocate resources towards public spending that prove difficult to reverse when the economic tide turns, leading to a "ratchet effect" (Hercowitz and Strawczynsk 2004). As these cyclical increases in public consumption persist, the size of the public sector steadily grows, subsequently limiting crucial fiscal space when economic downturns finally hit. In this paper, we explore "downward rigidity" on public consumption, which encompasses typically discretionary expenditures related to public employment (which includes wages, salaries and social contributions) and the procurement of goods and services. Adding to the predicament, emerging markets grapple with significant inefficiencies in public goods provision (Izquierdo, Pessino and Vuletin 2018). The rapid increase in public spending during booms can be an important source of these inefficiencies. Public finance and adjustment cost theory advises that desired increases in spending in health or education should be smooth rather than increase spasmodically over the business cycle. There are, broadly speaking, adjustment costs or time to build. Rapidly hiring teachers during an upturn when cash flushes may not permit careful vetting or drafting from successive graduate classes of training. Nor may there be time to learn from previous experience in building new medical centers. A growing number of studies precisely document the poor quality of much of public spending which, in turn, jeopardizes the future returns required to cover today’s investments (Dutu and Sicari 2016; Izquierdo, Pessino and Vuletin 2018; Joumard, et al. 2004; Miningou 2019). Ineffective spending on public education or health care, for instance, contributes to low labor productivity and diminishes the future economic returns necessary for repaying the initial expenditures. Therefore, the combination of inefficient and semi-procyclical (increasing in good times and remaining steady in bad times) public consumption sets the stage for fiscal stress even before the onset of a recession. Going beyond the well documented overall procyclicality found in emerging markets, this paper studies the asymmetric behavior of public consumption spending along their business cycle. While advanced economies are generally acyclical in their public consumption policies, we find emerging markets to be “semi-procyclical”. This means that they display a procyclical policy during booms and an acyclical policy during busts. We show that this asymmetry creates fiscal gaps as the expected positive correlation between spending and revenue breaks during economic downturns. 2. The Procyclical Puzzle: Emerging Markets and Fiscal Woes The fiscal behavior of emerging markets, characterized by procyclical tendencies, has been widely discussed in the public finance literature as the "original sin" responsible for their fiscal woes (see, e.g. Alesina, Campante and Tabellini 2008; Gavin and Perotti 1997; Ilzetzki and Végh 2008; Kaminsky, Reinhart and Végh 2005; Mendoza and Oviedo 2006; Talvi and Végh 2005; Tornall and Lane 1999). Contrary to standard Keynesian recommendations for stabilizing the economic cycle, empirical studies have revealed that overall primary spending in emerging markets shows a counterintuitive positive correlation with the state of the business cycle. This means that spending increases during good times and decreases during bad times. Updating previous research by Kaminsky, Reinhart and Végh (2005), and later by Frankel, Vegh and Vuletin (2013), Figure 1 illustrates the correlation between the cyclical component of general government primary expenditure (total government expenditure excluding interest payments) and the cyclical component of aggregate output (GDP) for 71 countries (22 advanced and 49 emerging) from 1980 to 2018. In the figure, red bars represent emerging economies, while blue bars represent advanced economies. A positive correlation between the cyclical components of primary government expenditure and GDP indicates procyclical fiscal policy, whereas a negative correlation indicates countercyclical policy. As shown, most advanced economies display countercyclicality, with exceptions such as Portugal and Spain, which exhibit spending patterns more akin to emerging markets. Conversely, emerging markets generally display procyclical fiscal policies. Figure 1: Correlation Between Cyclical Components of General Government Primary Expenditure and Gross Domestic Product Source: Authors’ calculations based on World Economic Outlook database 2019. Notes: The cyclical components have been estimated using the Hodrick-Prescott Filter with a lambda parameter equal to 6.25. The classification of emerging and advanced economies is provided in the appendix. This procyclicality has typically been attributed to "inherited initial conditions," such as political distortions, weak institutions, limited financial depth, and imperfect access to international credit markets (see, Caballero and Krishnamurthy 2004; Gavin and Perotti 1997; Gavin et al. 1996; Riascos and Végh 2003; Talvi and Végh 2005; Tornall and Lane 1999; Velasco 1999). Frankel, Végh and Vuletin (2013) show that institutional frameworks characterized by property rights protection, control of corruption, higher bureaucratic quality, and a strong law and order tradition, have enabled some developing countries to break free from procyclicality in the previous decade. Céspedes and Velasco (2014), using alternative proxies for institutional quality, find consistent evidence in a sample of 60 resource-rich countries. Additionally, Alesina, Campante and Tabellini (2008) establish a positive correlation between measures of corruption and procyclical fiscal policy. Procyclical public spending is a significant factor contributing to the excess economic volatility reported in emerging markets (Végh et al. 2018). Whereas too much economic volatility may foster economic uncertainty, hamper credit market development, and limit policy makers' ability to protect the country against negative shocks or harness positive ones,2 procyclical behavior, if symmetrical across the cycle, may not necessarily lead to fiscal unsustainability. Symmetric behavior along the business cycle implies that while bonanza cycles could be financed with increased public revenues and/or temporary borrowing, retrenchments would be sizeable enough to either adjust for lower revenue or cover those. Nonetheless, the recent history of emerging markets is plagued with episodes of fiscal turmoil. Much more than advanced economies.3 It is not just higher prevalence of fiscal trauma; the ensuing fiscal adjustments are also different in emerging markets. Emerging markets tend to address episodes of fiscal stress with desperate attempts at “fiscal reforms”, and disordered adjustment strategies which share little correspondence with those idealized in either the growth or equity literature.4 What gives in this puzzle? We argue that downwardly rigid spending creates fiscal gaps representing the seeds of later fiscal turmoil. 2 See, for example, World Bank (2013) and Baez, de la Fuente and Santos (2017). 3 Based on the Vulnerability Exercise for Emerging Market Economies (VEE) conducted by the International Monetary Fund (2010). Defining fiscal crisis periods as episodes of outright fiscal distress (public debt default/restructuring, need to access large-scale official/IMF support, hyperinflation) as well as extreme financing problems (spikes in sovereign bond spreads), Baldacci, Gupta and Mulas-Granados (2011) show how much more prevalent fiscal trauma is in emerging markets relative to advance economies. Indeed, since 1970, the authors estimated 135 episodes of fiscal stress in 52 emerging economies and just 41 in advanced economies. 4 While most fiscal stress episodes in advanced economies between 1970 and 2011 resulted in higher financing costs for relatively short periods of time, episodes in emerging markets led to longer and costlier adjustments, including 52 debt defaults or restructuring episodes and 79 IMF programs (Baldacci, Belhocine, et al. 2011). 3. Procyclicality of Public Consumption Meets Spending Rigidities Emerging markets face a unique challenge as discretionary public consumption exhibits downward rigidity during recessions. This arises from the difficulty of cutting back on public goods and services provision, and compensation to public employees. Emerging markets face an additional challenge as spending expansions often rely on increasing public wages or the number of employees, which have their own structural rigidities.5 While the overall cyclicality of public spending has been extensively studied, the asymmetrical responses of public consumption to the economic cycle have received limited attention. Among the few studies exploring these issues, Balassone and Francese (2004) find evidence in OECD countries of significant asymmetry in the reaction of fiscal policy to positive and negative cyclical conditions, with budgetary balances deteriorating in contractions and not improving in expansions. This asymmetry appears to have contributed significantly to debt accumulation. Similarly, Hercowitz and Strawczynsk (2004) find that the prolonged rise in the spending-to-output ratio is partially explained by cyclical upward ratcheting due to asymmetric fiscal behavior. These authors also analyzed cyclical changes in the composition of government spending, as well as a possible link between cyclical ratcheting and government weakness. For developing countries, Carneiro and Garrido (2016) investigate the extent to which countries behave procyclically or countercyclically in different phases of the business cycle, and find a causal link between stronger institutions and less procyclical fiscal policy, even after controlling for the endogeneity of institutions and other determinants of fiscal policy. Balassone and Kumar (2007) find evidence of exuberance in government expenditures during the boom phase of an economic cycle. They conclude that procyclicality may reflect an inaccurate assessment of the cycle, particularly in emerging markets during downturns. 5 Formally, we usually think of public spending rigidities as limits to modifying the level or structure of expenditure over a period imposed by institutional decrees (e.g., see Cetrángolo, Jiménez and Ruiz Del Castillo 2010; Echeverry, Bonilla and Moya 2006). The asymmetric cyclical behavior of public consumption is showcased in figure 2 below. In the figure, we compare the average growth rates of public spending on salaries, and goods and services, across the business cycle for both emerging and advanced economies.6 Figure 2: Average Growth in the Components of Government Consumption Over the Business Cycle Source: Authors’ calculations based on World Economic Outlook database 2019. Note: Each point represents the average growth of public consumption components (goods and services, and employee compensation) for each country in the sample. Red dots indicate averages during recessions, while blue dots represent averages during expansions. We observe that, on average, emerging economies exhibit significantly higher growth rates in both spending categories compared to advanced economies. This contrast is especially pronounced during periods of expansion, where emerging economies typically increase expenditure by a 6 Recessions are country specific and are obtained from a transition function for each country that ranges between 0 (largest expansion) and 1 (deepest recession). We define recessions as values where ≥ 0.5 and expansions those episodes where < 0.5. See econometric section for details on this function. maximum of 15 percent, while advanced economies limit their spending growth to a maximum of 5 percent on average. Furthermore, when analyzing the behavior of both groups during recessions, the differences remain evident. Advanced economies tend to experience near-zero growth rates during economic downturns, whereas emerging economies present growth rates ranging from minus 5 to positive 10 percent. This highlights the asymmetry in public consumption within emerging economies, where the substantial overspending during expansions is not matched by a proportional retrenchment during contractions. In contrast, advanced economies exhibit more consistent spending behavior throughout the economic cycle. Public consumption expenditures play a crucial role in facilitating the provision of essential public goods such as education, health care, law enforcement and defense. Therefore, it is not apparent why governments would adjust the number of teachers, doctors, police officers, or military personnel in response to fluctuations in the business cycle. In theory, these decisions should be driven by societal preferences and the unique characteristics of the public goods production function. In fact, that is what we observe among advanced economies, where growth of public spending appears to be very similar throughout varying business cycles. The only visible changes over time should reflect the long-term trend of the increasing share of the public sector in the national product as countries develop (Wagner 1958). However, emerging markets’ behavior looks quite different. Emerging economies are procyclical in good times, but those are not followed by a retrenchment that is enough to cover the overspending during expansion. Once again, political economy arguments offer insights into why certain expenditures exhibit stronger cyclical responses compared to others, shedding light on the underlying dynamics of public spending. While public spending rigidities are commonly associated with institutional constraints, such as contractual obligations, entitlements, and interest payments; it can also be subject to other serious forms of rigidities. Institutional weaknesses and political economy factors can compromise the ability or willingness of governments to apply discretionary cuts to public consumption and the provision of essential public goods. Tornall and Lane (1999) developed a model where several political groups compete for fiscal resources. Using the well-known consequences of the problem of the common pool (Ostrom 2015), the authors show how this competition leads to a “voracity effect”, defined as a disproportioned response of public spending to exogenous shocks in the economy, such as unexpected windfalls from commodity exports. Similarly, Talvi and Végh (2005) proposed a model where political actors increase the pressure directed to public spending during periods of economic bonanza. Meanwhile, Lane (2003) shows how procyclical tendencies of emerging countries may grow with the number of political actors with distinct goals and constituencies. Divergent political views and policy objectives among political groups may also contribute to overspending during economic booms (e.g., see Humphreys and Sandbu 2007; and Ilzetzki 2011). This downward rigidity in public consumption makes emerging markets "semi-procyclical" rather than fully procyclical. This behavior not only subjects the size of the public sector to cyclical fluctuations, but also prevents these economies from effectively self-insuring against fiscal limitations during downturns. This is particularly costly for economies with significant tax base volatility and limited access to international credit during recessions (Gavin et. al 1996; Riascos and Végh 2003; Talvi and Végh 2005). To illustrate these cases with specific examples, in Figure 3, panels a and b depict the evolution of various categories in general government total expenditure as a percent of GDP. Argentina serves as a prime example of the issue we are describing, especially concerning the employee category. Figure 3. Evolution of Total Government Expenditure for Argentina and United States a. Panel A b. Panel B Source: Authors’ calculations based on World Economic Outlook database 2019. Note: The ratios are calculated using the seasonally adjusted real GDP from the database. The transformation is performed using the Hodrick-Prescott filter with a lambda equal to 6.25. This approach is used to avoid cyclical components in the analysis. The recession periods are defined using the criteria from Auerbach and Gorodnichenko (2011), employing the same variable rec described in the econometric section, but using the deviation of the output growth rate from its trend for each year. Moreover, contrary to the rest of the paper, we extract the trend using the Hodrick-Prescott filter with a very high smoothing parameter (lambda = 10,000) so that the trend is very smooth. We then normalize the deviation for each country and assign it to the rec variable, defining a recession as values higher than 0.7. In this way, we can capture variations in the usual growth rate for each country and identify the exact year of contraction. The marked recessions for the United States coincide with those published by the Federal Reserve. We can observe that during the bonanza times due to the commodity super-cycle of the 2000s, Argentina increased the share of GDP dedicated to public employee expenditure from 7.4 percent in 2002 to 10.6 percent in 2009. Once it reached that level, it did not increase significantly in the following years; however, the size in the GDP never reverted to the pre-expansion period. On the other hand, the United States presents a different picture. While the components of expenditure remain practically constant during years of expansion, increases occur during crises only to be contracted afterward, except for the case of social benefits, which, after the 2009 crisis, never returned to their original level. As a result, the United States has the capacity to perform countercyclical fiscal policy without risking financial stability, as it is able to expand and then contract to achieve fiscal soundness. In contrast, Argentina experiences a fiscal frenzy during good times that is not followed by retrenchment or mild reduction, but rather a continuation of the ascending trend, or at best, a plateau. In this way, over ten years, Argentina's spending on social benefits increased from 14 percent of GDP in 2008 to 18 percent in 2018. 4. Data and Estimation We analyze asymmetric cyclical behavior in emerging and advanced economies using the IMF's World Economic Outlook dataset from April 2019. We chose this dataset over newer versions due to its detailed disaggregation of government expenditure, which is no longer available in more recent publications. The frequency of the data is annual. The panel dataset spans from up to 1980 to 2018 and includes 71 countries from various regions and income classifications. Our findings contrast evidence from up to 49 emerging economies and 22 advanced countries. A comprehensive list of these countries is provided in Table A1 in the appendix. We employ a non-linear econometric model to estimate the elasticities of public consumption relative to aggregate output throughout the business cycle. Table 1 presents descriptive statistics for the growth rates of the specific categories of public spending under analysis, as well as the real GDP used in our regressions. Table 1: Descriptive Statistics Advanced Economies Variable Obs Mean Std. Dev. Min Max GDP Growth 836 .024 .022 -.086 .126 Compensation to Employees 469 .018 .028 -.166 .129 Growth Goods and Services Growth 471 .023 .039 -.16 .255 Emerging Markets Variable Obs Mean Std. Dev. Min Max GDP Growth 1740 .038 .049 -.548 .368 Compensation to Employees 922 .044 .097 -1.234 .748 Growth Goods and Services Growth 861 .04 .172 -.914 1.411 Source: Authors’ calculations based on World Economic Outlook database 2019. We can observe that, regarding GDP growth, while emerging markets show a higher average rate for the period considered, they also exhibit greater dispersion. This observation is consistent with existing literature (Végh et al. 2018), which suggests that emerging markets exhibit more pronounced economic fluctuations, experiencing both more severe downturns and more robust upswings compared to advanced economies, which typically demonstrate lower but more stable growth rates over time. Similarly, for compensation to employees, we observe the same pattern: the growth rate in this category is 2.4 times higher in emerging markets compared to advanced economies, and this higher rate is associated to greater variability. Additionally, on average, the growth of the goods and services category is 1.7 times higher in emerging markets than in advanced economies, and this category demonstrates more dispersion. 5. Identification Strategy Our primary economic question focuses on how public spending reacts to changes in income, specifically analyzing the cyclicality of spending. A major concern in estimating this relationship is the potential endogeneity due to reverse causality between public consumption and aggregate income. The literature on spending multipliers extensively argues that public spending can influence aggregate output, making it challenging to disentangle the direction of causality in our estimates (e.g., Barnichon, Debortoli and Matthes 2022; Christiano, Eichenbaum and Rebelo 2011; Kraay 2012). To address the reverse causality bias in our cyclicality estimates, we adopt an instrumental variables (IV) approach. This allows us to isolate exogenous variations in income that are not directly influenced by changes in public spending, thus providing a more accurate estimate of the cyclicality of public expenditure. Our IV strategy hinges on a two-step procedure. First, we argue that domestic public consumption policy likely has limited direct effects on the aggregate income of major trade partners. However, sudden income changes in these trade partners can have substantial impacts on domestic output. This assumption forms the basis of our exclusion restriction: while shocks from the incomes of major trade partners are plausibly exogenous to domestic public consumption policy, they still exert significant influence on domestic income. Therefore, using trade partner income shocks as an instrument for domestic income should satisfy the exclusion restriction, allowing us to estimate the impact of income changes on public spending without the bias of reverse causality. Recent studies have used idiosyncratic variation in trade partner’s income as exogenous instruments to study domestic conditions, Cravino and Levchenko (2017) examine how multinational firms play a role in the transmission of international business cycle shocks across countries, making use of global income variations as exogenous instruments. This provides a framework for understanding how foreign income shocks can affect domestic outcomes. Di Giovanni, Levchenko and Mejean (2014) explore how aggregate fluctuations are influenced by trade partners and specific destinations, using foreign demand shocks as an instrument to analyze fluctuations in output. Their approach would be quite relevant to our strategy of using exogenous income shocks from trade partners. More recently, a study by Lloyd, Manuel and Panchev (2024) explores the role of global drivers, such as foreign-weighted GDP, in shaping domestic economic risks. The first stage in our two step-least squares dummy variable procedure follows: , = + 1 , ∗ (1 − , ) + 2 , ∗ , + ,−1 + , (1) Where , is a vector of the domestic real GDP growth rate for country i and time t, , represents the trade-weighted growth rate of real output in trade partners, and , is a country-specific continuous transition function ranging between 0 (largest expansion) and 1 (deepest recession). The trade-weighted measure is defined as: 10 , = � ,, ∗ , =1 , represents the instrument for country i in year t. The variable , denotes the export share of country i to its trading partner j in year t, while , captures the real output growth of trading partner j in year t. The top 10 trading partners are selected for each country i based on export shares, with a minimum threshold of 30 percent of total exports to ensure that the selected partners account for a significant portion of the country’s trade. Thus, the instrument reflects the trade- weighted average of the economic growth of the most relevant trading partners, effectively capturing external demand. The key assumption is that the growth of trading partners primarily influences the GDP growth of country i, with limited or no direct effect on domestic public spending decisions through other channels. Drawing from Auerbach and Gorodnichenko (2011), our proxy function for the economic cycle is defined as: −, , = 1+ −, (2) Here, , is a normalized variable that measures the state of the business cycle. It is calculated using a 3-year moving average of the growth rate of real output and then normalized so that ( ) = 0 and ( ) = 1 for each country i. Finally, ,−1 is a set of controls including the lag of total public expenditure and the lag the dependent variable. The second stage follows: � � , = + 1 , ∗ (1 − , ) + 2 , ∗ , + ,−1 + , (3) Where , is a vector of the real growth rate of public spending components for country i and time �, represents the estimated growth rate of real output in the first stage. Here 1 t, and 2 represent the causal effect of income to public spending during expansions and recessions respectively. To uncover differences across different country groups, we split the sample between advanced and emerging economies. Table 2. Changes in GDP Growth Across the Business Cycle on Public Consumption: Elasticities from Fixed Effects and IV Regressions for Emerging and Advanced Economies (1) (2) (3) (4) VARIABLES Emerging Advanced Emerging Advanced economies - economies - economies - economies - FE FE IV IV GDP Growth * Exp 1.12 0.37 1.10 0.40 [0.136]*** [0.109]*** [0.143]*** [0.091]*** GDP Growth * Rec 0.43 -0.22 0.36 -0.24 [0.244]* [0.129] [0.249] [0.141]* Lagged Public Consumption Growth -0.05 0.30 -0.05 0.30 [0.054] [0.096]*** [0.054] [0.102]*** Lagged General Government Expense Growth 0.05 0.04 0.05 0.04 [0.034] [0.065] [0.047] [0.058] Constant 0.01 0.01 [0.006]* [0.003]*** Observations 800 432 793 418 R-squared 0.122 0.158 Number of countries 39 18 39 18 Country Fixed Effect Yes Yes Yes Yes Adjusted R-squared 0.071 0.112 First Stage IV Estimates First Stage First Stage Trade Weighted Growth * Exp 1.17 1.27 [0.03]*** [0.02]*** Trade Weighted Growth * Rec 1.35 1.43 [0.03]*** [0.04]*** Underidentification - Weak Identification - Weak Instruments Tests LM test statistic for underidentification 84.20 33.38 p-value of underidentification LM statistic (Anderson or Kleibergen-Paap) 0 7.58e-09 F statistic for weak identification (Cragg-Donald or Kleibergen-Paap) 918 813.8 Anderson-Rubin chi-sqared test of significance of endogenous regressors 65.43 22.66 p-value of Anderson-Rubin chi-sqared test of endogenous regressors 0 1.20e-05 Source: Authors’ calculations based on World Economic Outlook database 2019. Notes: Robust standard errors in brackets. *** p<0.01, ** p<0.05, * p<0.1 Underidentification tests. Ho: matrix of reduced form coefficients has rank=K1-1 -underidentified. Ha: matrix has rank=K1-identified. Weak identification test. Ho: equation is weakly identified. Weak-instrument-robust inference. Tests of joint significance of endogenous regressors B1 in main equation. Ho: B1=0 and overidentifying restrictions are valid. Table 2 presents the results using a fixed effects estimator (Columns 1 and 2) and an instrumental variable estimator (Columns 3 and 4). During periods of expansion (when approaches zero and approaches one), changes in output are associated with a more than proportional increase in public consumption in developing economies. In contrast, the same output change leads to a less than proportional effect in advanced economies. The semi-procyclical behavior is visible from the coefficient on output growth during recessions, particularly in emerging economies, where spending is not reduced during downturns. This asymmetrical response results in the downward rigidity previously highlighted. These findings are consistent across both the fixed effects and instrumental variable estimators. The null hypothesis of the Kleibergen-Paap LM test is that the structural equation is underidentified, meaning the rank condition fails. This test applies the rank test procedure introduced by Kleibergen and Paap (2006). Given that the associated p-values are below 1 percent, we can confidently reject the null hypothesis of underidentification for both IV regressions. Additionally, the weak identification test confirms that the instruments are strong, as the F- statistics far exceed the threshold of 10, indicating no issues with weak identification. Finally, we reject the null hypothesis that the endogenous variables are irrelevant or that their coefficients are zero, supported by the sufficiently low p-values. Taking advantage of the continuous nature of our business cycle proxy, we use the estimates from our IV specification to visualize the output elasticities of public consumption across varying degrees of economic recessions and expansions. Figure 4 below depicts the output elasticities at different stages of the business cycle for advanced economies and emerging markets. Figure 4: Output Elasticity of Public Consumption Over the Business Cycle Source: Authors’ calculations based on World Economic Outlook database 2019. According to Figure 4, advanced economies tend to maintain moderate spending during economic expansions, with retrenchment during downturns being statistically insignificant. Conversely, spending decisions in emerging markets are significantly more responsive to short-term economic conditions. In these economies, periods of growth lead to substantial increases in components of public consumption. In opposition, contractions do not result in reductions in spending, suggesting a form of downward spending rigidity and contributing to the semi-procyclicality of public consumption. 6. From Semi-Procyclicality to Fiscal Gaps An important part of our story is that the spasmodic and quite common fiscal turmoil experienced by emerging markets may well have its roots during good times. One of the most important consequences of semi-procyclical behaviors is that the lack of symmetrical retrenchment as the cycle turns open partial deficits that need to be financed either through debt or through compositional changes in spending. To be able to sustain the previous statement, we need to incorporate into our analysis the relationship between public consumption and revenue along the cycle. We do not use overall deficits to capture partial financing needs as some countries use change in spending composition to close these financing gaps instead of debt. The literature has identified large biases against public investment used to cover fiscal holes in other spending categories (Ardanaz and Izquierdo 2017). Biases against public investment could be as problematic for economic growth as large increases in public debt (Calderón, Easterly and Servén 2003). Our identification strategy follows closely the IV methodology used in our main section. Here, we again use exogenous income shocks provided by trade-weighted changes in foreign demand to predict total primary revenue. Once we find exogenous changes in revenue, we look at their correlation with changes in spending policy. Thus, the asymmetries along the cycle in the correlations between public spending and revenue are captured using the following two-step IV model: 1 : , = + 1 , ∗ (1 − , ) + 2 , ∗ , + ,−1 + , (4) � � 2 : , = + 1 , ∗ �1 − , � + 2 , ∗ , + ,−1 + , (5) Here , represent total public revenue growth and we use the same vector of lagged controls ,−1 as in our previous analysis. Table 3. Changes in Revenue Growth Across the Business Cycle on Public Consumption: Elasticities from Fixed Effects and IV Regressions for Emerging and Advanced Economies (1) (2) (3) (4) VARIABLES Emerging Advanced Emerging Advanced economies - economies - economies - economies - FE FE IV IV Revenue Growth * Exp 0.31 0.37 0.89 0.40 [0.104]*** [0.081]*** [0.177]*** [0.086]*** Revenue Growth * Rec -0.11 -0.01 0.09 -0.27 [0.169] [0.098] [0.241] [0.153]* Lagged Public Consumption Growth -0.02 0.31 -0.07 0.31 [0.060] [0.094]*** [0.062] [0.115]*** Lagged General Government Expense Growth 0.05 0.02 0.06 0.01 [0.034] [0.062] [0.058] [0.057] Constant 0.04 0.01 [0.005]*** [0.002]*** Observations 791 432 784 418 R-squared 0.050 0.190 Number of countries 39 18 39 18 Country Fixed Effect Yes Yes Yes Yes Adjusted R-squared -0.337 0.108 First Stage IV Estimates First Stage First Stage Trade Weighted Growth * Exp 1.43 1.32 [.21]*** [.08]*** Trade Weighted Growth * Rec 1.72 1.39 [.29]*** [.15]*** Underidentification - Weak Identification - Weak Instruments Tests LM test statistic for underidentification 41.87 30.31 p-value of underidentification LM statistic (Anderson or Kleibergen-Paap) 9.77e-11 3.68e-08 F statistic for weak identification (Cragg-Donald or Kleibergen-Paap) 21.85 51.14 Anderson-Rubin chi-sqared test of significance of endogenous regressors 64.05 22.66 p-value of Anderson-Rubin chi-sqared test of endogenous regressors 0 1.20e-05 Source: Authors’ calculations based on World Economic Outlook database 2019. Notes: Robust standard errors in brackets. *** p<0.01, ** p<0.05, * p<0.1 Underidentification tests. Ho: matrix of reduced form coefficients has rank=K1-1 -underidentified. Ha: matrix has rank=K1-identified. Weak identification test. Ho: equation is weakly identified. Weak-instrument-robust inference. Tests of joint significance of endogenous regressors B1 in main equation. Ho: B1=0 and overidentifying restrictions are valid. Table 3 shows the results of running this analysis for our sample of emerging and advanced economies. Column (3) looks at the correlation of public consumption to primary revenue and shows how spending grows at more or less the same rate as revenue (beta estimates are very close to one) during good times. But as the cycle turns and revenues start to decline (shown as a positive and symmetric relationship between foreign income changes and revenue), the correlation breaks down and spending becomes unresponsive to the declining revenue. While the beta estimates are similar and statistically close to 1 for both components of public consumption, salaries and goods and services, it looks like the latter tends to overreact to revenue relative to the former (Table A4 and Table A5 in the appendix). This makes sense as the rigid nature of public salaries may prevent quick adjustments as revenue climbs. There may be no such impediments to adjusting spending on goods and services as new revenue starts to come in. To conclude, the asymmetric correlation between public consumption and primary revenue shows that the semi-procyclical behavior discussed above leads to partial deficits in need of financing. Using debt or decreases in public investment to cover those fiscal holes may carry significant adverse growth effects to emerging markets. 7. Conclusions This paper studies the dynamics of public consumption within emerging and advanced economies, focusing on the asymmetric responses to the business cycle. Our analysis reveals significant differences in fiscal behavior between the two groups, with implications for fiscal policy. The results display that in advanced economies, public consumption experiences moderate growth during economic expansions. Conversely, in emerging economies, output growth induces a more pronounced positive effect on both components of public consumption, which becomes acyclical during recessions. Moreover, the elasticity observed in goods and services for both groups diminishes when employee compensation is the dependent variable, highlighting the rigidities inherent in this category. The policy implications are that, in order to mitigate the procyclical and inefficient nature of public consumption, emerging markets need to enhance their fiscal frameworks. Improved medium-term fiscal planning, overseen by independent fiscal institutions, can provide the necessary oversight. Incorporating fiscal rules can help stabilize public spending and reduce fiscal vulnerabilities. In conclusion, achieving fiscal sustainability in emerging markets requires a robust fiscal framework, effective institutional oversight, and targeted reforms. By adopting these measures, emerging economies can better navigate the cyclical nature of their economies, ensuring more sustainable public finances in the long term. 8. Appendix Table A1. Classification of Countries Included in the Analysis Advanced Economies Australia Japan Austria Luxembourg Belgium Netherlands Canada New Zealand Denmark Norway Finland Portugal France Spain Germany Sweden Hong Kong SAR, China Switzerland Israel United Kingdom Italy United States Emerging Markets Argentina Mauritius Bahrain, Kingdom of Mexico Bangladesh Morocco Botswana Namibia Brazil Nigeria Bulgaria Pakistan Chile Panama China, P.R.: Mainland Peru Colombia Philippines Croatia Poland Cyprus Qatar Czechia Romania Côte d'Ivoire Russian Federation Egypt, Arab Rep. Slovak Republic Estonia Slovenia Ghana South Africa Hungary Sri Lanka Iceland Taiwan, China India Thailand Indonesia Tunisia Jamaica Türkiye Kazakhstan United Arab Emirates Lebanon Viet Nam Lithuania Zambia Malaysia Table A2. Changes in GDP Growth Across the Business Cycle on Compensation to Employees: Elasticities from Fixed Effects and IV Regressions for Emerging and Advanced Economies (1) (2) (3) (4) VARIABLES Emerging Advanced Emerging Advanced economies - economies - economies - economies - FE FE IV IV GDP Growth * Exp 0.92 0.34 0.92 0.34 [0.145]*** [0.124]** [0.159]*** [0.095]*** GDP Growth * Rec 0.34 -0.15 0.29 -0.16 [0.218] [0.125] [0.237] [0.145] Lagged Compensation to Employees Growth -0.06 0.26 -0.06 0.29 [0.088] [0.066]*** [0.084] [0.122]** Lagged General Government Expense Growth 0.07 0.07 0.07 0.03 [0.039]* [0.086] [0.045] [0.059] Constant 0.02 0.01 [0.006]** [0.003]** Observations 873 432 866 418 R-squared 0.086 0.128 Number of Countries 42 18 42 18 Country Fixed Effect Yes Yes Yes Yes Adjusted R-squared 0.034 0.082 First Stage IV Estimates First Stage First Stage Trade Weighted Growth * Exp 1.19 1.27 [.03]*** [.02]*** Trade Weighted Growth * Rec 1.36 1.43 [.03]*** [.04]*** Underidentification - Weak Identification - Weak Instruments Tests LM test statistic for underidentification 84.54 33.42 p-value of underidentification LM statistic (Anderson or Kleibergen-Paap) 0 7.42e-09 F statistic for weak identification (Cragg-Donald or Kleibergen-Paap) 857.5 804.6 Anderson-Rubin chi-sqared test of significance of endogenous regressors 40.41 13.91 p-value of Anderson-Rubin chi-sqared test of endogenous regressors 1.68e-09 0.000952 Source: Authors’ calculations based on World Economic Outlook database 2019. Robust standard errors in brackets.*** p<0.01, ** p<0.05, * p<0.1.Underidentification tests. Ho: matrix of reduced form coefficients has rank=K1-1 -underidentified. Ha: matrix has rank=K1-identified. Weak identification test. Ho: equation is weakly identified. Weak-instrument-robust inference. Tests of joint significance of endogenous regressors B1 in main equation. Ho: B1=0 and overidentifying restrictions are valid. Table A3. Changes in GDP Growth Across the Business Cycle on Goods and Services: Elasticities from Fixed Effects and IV Regressions for Emerging and Advanced Economies (1) (2) (3) (4) VARIABLES Emerging Advanced Emerging Advanced economies - economies - economies - economies - FE FE IV IV Real GDP Growth * Exp 1.60 0.48 1.58 0.63 [0.218]*** [0.183]** [0.526]*** [0.139]*** Real GDP Growth * Rec 0.38 -0.36 0.29 -0.45 [0.366] [0.214] [0.399] [0.166]*** Lagged Goods and Services Growth -0.22 0.12 -0.22 0.07 [0.095]** [0.092] [0.070]*** [0.060] Lagged General Government Expense Growth 0.02 0.13 0.01 0.19 [0.058] [0.079] [0.087] [0.086]** Constant 0.00 0.01 [0.009] [0.004]*** Observations 817 432 810 418 R-squared 0.089 0.078 Number of Countries 40 18 40 18 Country Fixed Effect Yes Yes Yes Yes Adjusted R-squared 0.036 0.052 First Stage IV Estimates First Stage First Stage Trade Weighted Growth * Exp 1.18 1.27 [.03]*** [.02]*** Trade Weighted Growth * Rec 1.35 1.43 [.03]*** [.04]*** Underidentification - Weak Identification - Weak Instruments Tests LM test statistic for underidentification 92.65 35.91 p-value of underidentification LM statistic (Anderson or Kleibergen-Paap) 0 2.07e-09 F statistic for weak identification (Cragg-Donald or Kleibergen-Paap) 1008 856.6 Anderson-Rubin chi-sqared test of significance of endogenous regressors 15.59 28.69 p-value of Anderson-Rubin chi-sqared test of endogenous regressors 0.000411 5.89e-07 Source: Authors’ calculations based on World Economic Outlook database 2019. Robust standard errors in brackets.*** p<0.01, ** p<0.05, * p<0.1.Underidentification tests. Ho: matrix of reduced form coefficients has rank=K1-1 -underidentified. Ha: matrix has rank=K1-identified. Weak identification test. Ho: equation is weakly identified. Weak-instrument-robust inference. Tests of joint significance of endogenous regressors B1 in main equation. Ho: B1=0 and overidentifying restrictions are valid. Table A4. Changes in Revenue Growth Across the Business Cycle on Compensation to Employees: Elasticities from Fixed Effects and IV Regressions for Emerging and Advanced Economies (1) (2) (3) (4) VARIABLES Emerging Advanced Emerging Advanced economies - economies - economies - economies - FE FE IV IV Revenue Growth * Exp 0.28 0.32 0.74 0.34 [0.121]** [0.094]*** [0.167]*** [0.090]*** Revenue Growth * Rec -0.07 0.06 0.10 -0.18 [0.149] [0.110] [0.220] [0.152] Lagged Compensation to Employees Growth -0.04 0.26 -0.06 0.30 [0.093] [0.068]*** [0.083] [0.130]** Lagged General Government Expense Growth 0.09 0.05 0.08 0.01 [0.039]** [0.079] [0.051] [0.056] Constant 0.04 0.01 [0.004]*** [0.001]*** Observations 864 432 857 418 R-squared 0.045 0.162 Number of Countries 42 18 42 18 Country Fixed Effect Yes Yes Yes Yes Adjusted R-squared -0.200 0.090 First Stage IV Estimates First Stage First Stage Trade Weighted Growth * Exp 1.45 1.3 [.2]*** [.08]*** Trade Weighted Growth * Rec 1.68 1.39 [.24]*** [.15]*** Underidentification - Weak Identification - Weak Instruments Tests LM test statistic for underidentification 43.97 30.51 p-value of underidentification LM statistic (Anderson or Kleibergen-Paap) 0 3.33e-08 F statistic for weak identification (Cragg-Donald or Kleibergen-Paap) 29.28 52.01 Anderson-Rubin chi-sqared test of significance of endogenous regressors 39.88 13.91 p-value of Anderson-Rubin chi-sqared test of endogenous regressors 2.19e-09 0.000952 Source: Authors’ calculations based on World Economic Outlook database 2019. Robust standard errors in brackets.*** p<0.01, ** p<0.05, * p<0.1.Underidentification tests. Ho: matrix of reduced form coefficients has rank=K1-1 -underidentified. Ha: matrix has rank=K1-identified. Weak identification test. Ho: equation is weakly identified. Weak-instrument-robust inference. Tests of joint significance of endogenous regressors B1 in main equation. Ho: B1=0 and overidentifying restrictions are valid. Table A5. Changes in Revenue Growth Across the Business Cycle on Goods and Services: Elasticities from Fixed Effects and IV Regressions for Emerging and Advanced Economies (1) (2) (3) (4) VARIABLES Emerging Advanced Emerging Advanced economies - economies - economies - economies - FE FE IV IV Revenue Growth * Exp 0.37 0.47 1.32 0.63 [0.106]*** [0.108]*** [0.484]*** [0.133]*** Revenue Growth * Rec -0.22 -0.18 -0.06 -0.50 [0.164] [0.099]* [0.397] [0.189]*** Lagged Goods and Services Growth -0.20 0.15 -0.24 0.09 [0.102]* [0.086] [0.076]*** [0.061] Lagged General Government Expense Growth 0.03 0.09 -0.02 0.14 [0.060] [0.074] [0.100] [0.086] Constant 0.04 0.01 [0.006]*** [0.002]*** Observations 808 432 801 418 R-squared 0.050 0.092 Number of Countries 40 18 40 18 Country Fixed Effect Yes Yes Yes Yes Adjusted R-squared -0.202 0.028 First Stage IV Estimates First Stage First Stage Trade Weighted Growth * Exp 1.41 1.33 [.2]*** [.08]*** Trade Weighted Growth * Rec 1.77 1.36 [.29]*** [.15]*** Underidentification - Weak Identification - Weak Instruments Tests LM test statistic for underidentification 44.93 29.74 p-value of underidentification LM statistic (Anderson or Kleibergen-Paap) 0 4.95e-08 F statistic for weak identification (Cragg-Donald or Kleibergen-Paap) 23.97 47.36 Anderson-Rubin chi-sqared test of significance of endogenous regressors 15.45 28.69 p-value of Anderson-Rubin chi-sqared test of endogenous regressors 0.000442 5.89e-07 Source: Authors’ calculations based on World Economic Outlook database 2019. Robust standard errors in brackets.*** p<0.01, ** p<0.05, * p<0.1.Underidentification tests. Ho: matrix of reduced form coefficients has rank=K1-1 -underidentified. Ha: matrix has rank=K1-identified. Weak identification test. Ho: equation is weakly identified. Weak-instrument-robust inference. Tests of joint significance of endogenous regressors B1 in main equation. Ho: B1=0 and overidentifying restrictions are valid. Figure A1. Output Elasticities of Public Consumption Components Over the Business Cycle Panel A Panel B Source: Authors’ calculations based on World Economic Outlook database 2019. 6. References Alesina, Alberto, Filipe R. Campante, and Guido Tabellini. 2008. "Why is Fiscal Policy Often Procyclical?" Journal of the European Economic Association 1006–1036. Ardanaz, Martín, and Alejandro Izquierdo. 2017. "Current expenditure upswings in good times and capital expenditure downswings in bad times?: New evidence from developing countries." IDB Working Paper (IDB Working Paper No. IDB-WP-838) No. IDB-WP- 838. Auerbach, Alan J., and Yuriy Gorodnichenko. 2011. "Fiscal Multipliers in Recession and Expansion." National Bureau of Economic Research no. w17447. Baez, Javier E, Alejandro de la Fuente, and Indhira Santos. 2017. "Do Natural Disasters Affect Human Capital? An Assessment Based on Existing Empirical Evidence." The World Bank Research Observer 33(2): 235–263. doi:10.1093/wbro/lkx006. Balassone, Fabrizio, and Manmohan Kumar. 2007. "Cyclicality of fiscal policy." In Promoting fiscal discipline, by Manmohan S. Kumar and Teresa Ter-Minassian, 19-35. Washington D.C.: International Monetary Fund. Balassone, Fabrizio, and Maura Francese. 2004. Cyclical Asymmetry in Fiscal Policy, Debt Accumulation and the Treaty of Maastricht. Rome: Bank of Italy. Baldacci, Emanuele, Nazim Belhocine, Gabriela Dobrescu, and Iva Petrova. 2011. "Assessing Fiscal Stress." IMF Working Papers (IMF Working Papers) (Working Paper No. 2011/100): 43. Baldacci, Emanuele, Sanjeev Gupta, and Carlos Mulas-Granados. 2011. "Fiscal Crises, Causes, Consequences, and Policy Responses." IMF Working Paper WP/11/280. Barnichon, Regis, Davide Debortoli, and Christian Matthes. 2022. "Understanding the size of the government spending multiplier: It’s in the sign." The Review of Economic Studies 89.1: 87-117. Caballero, Ricardo J., and Arvind Krishnamurthy. 2004. "Fiscal Policy and Financial Depth." National Bureau of Economic Research Working Paper 10532. Calderón, César, William Easterly, and Luis Servén. 2003. "Latin America’s infrastructure in the era of macroeconomic crises." In The Limits of Stabilization, Infrastructure, Public Deficits and Growth in Latin America, by William Easterly and Luis Servén, 21-94. Palo Alto, CA: Stanford University and Washington, DC: World Bank. Carneiro, Francisco, and Leonardo Garrido. 2016. "New Evidence on the Cyclicality of Fiscal Policy." World Bank Policy Research Working Paper No. 7293. Céspedes, Luis Felipe, and Andrés Velasco. 2014. "Was this time different?: Fiscal policy in commodity republics." Journal of Development Economics 92-106. Cetrángolo, Oscar, Juan Pablo Jiménez, and Ramiro Ruiz Del Castillo. 2010. Rigidities and Fiscal Space in Latin America: a Comparative Case Study. Serie Macroeconomía del Desarrollo, Santiago: Economic Comission for Latin America and the Caribbean (ECLAC). Christiano, Lawrence, Martin Eichenbaum, and Sergio Rebelo. 2011. "When is the government spending multiplier large?" Journal of Political Economy 119.1: 78-121. Cravino, Javier, and Andrei A. Levchenko. 2016. "Multinational Firms and International Business Cycle Transmission." National Bureau of Economic Research no. w22498. Cravino, Javier, and Andrei A. Levchenko. 2017. "Multinational Firms and International Business Cycle Transmission." The Quarterly Journal of Economics, President and Fellows of Harvard College, vol. 132(2) 921-962. Di Giovanni, Julian, Andrei A. Levchenko, and Isabelle Mejean. 2014. "Firms, Destinations, and Aggregate Fluctuations." Econometrica, Vol. 82, No. 4 (July, 2014) 1303-1340. Dutu, Richard, and Patrizio Sicari. 2016. Public Spending Efficiency in the OECD: Benchmarking Health Care, Education and General Administration. OECD Publishing, Paris: OECD Economics Department Working Papers, No. 1278. Echeverry, Juan Carlos, Jorge Alexander Bonilla, and Andrés Moya. 2006. Institutional Rigidities and Budget Flexibility: Origin, Motivation, and Effects on the Budget. Public Policy Management and Transparency Network: Effectiveness in Development and Result-based Budgetary Management, Washington: Inter-American Development Bank (IADB). Frankel, Jeffrey A., Carlos A. Vegh, and Guillermo Vuletin. 2013. "On graduation from fiscal procyclicality." Journal of Development Economics 32-47. Gavin, Michael, and Roberto Perotti. 1997. "Fiscal Policy in Latin America." NBER Macroeconomics Annual Volume 12 11-61. Gavin, Michael, Ricardo Hausmann, Roberto Perotti, and Ernesto Talvi. 1996. "Managing Fiscal Policy in Latin America and the Caribbean: Volatility, Procyclicality, and Limited Creditworthiness." IDB Working Paper No. 269. Hercowitz, Zvi, and Michel Strawczynsk. 2004. " Cyclical Ratcheting in Government Spending: Evidence from the OECD." The Review of Economics and Statistics 353–361. Humphreys, Macartan, and Martin E. Sandbu. 2007. "The Political Economy of Natural Resource Funds." In Escaping the Resource Curse, by Macartan Humphreys, Jeffrey D. Sachs and Joseph E. Stiglitz, 194-233. New York: Columbia University Press. Ilzetzki, Ethan. 2011. "Rent-seeking distortions and fiscal procyclicality." Journal of Development Economics 30-46. Ilzetzki, Ethan, and Carlos A. Vegh. 2008. "Procyclical Fiscal Policy in Developing Countries: Truth or Fiction?" NBER Working Paper No. 14191. International Monetary Fund. 2010. The IMF’s Vulnerability Exercise for Emerging Economies. Washington, DC: International Monetary Fund. Izquierdo, Alejandro, Carola Pessino, and Guillermo Vuletin. 2018. Better spending for better lives: how Latin America and the Caribbean can do more with less. Washington: Inter- American Development Bank. Joumard, Isabelle, Per Mathis Kongsrud, Young-Sook Nam, and Robert Price. 2004. Enhancing the Effectiveness of Public Spending: Experience in OECD Countries. OECD Economics Department Working Papers, No. 380, Paris: OECD Publishing. Kaminsky, Graciela L., Carmen M. Reinhart, and Carlos A. Végh. 2005. "When it Rains, it Pours: Procyclical Capital Flows and Macroeconomic Policies." NBER Macroeconomics Annual 2004 Volume 19. Kleibergen, Frank, and Richard Paap. 2006. "Generalized reduced rank tests using the singular value decomposition." Journal of econometrics, vol. 133, no 1 97-126. Kraay, Aart. 2012. "How large is the government spending multiplier? Evidence from World Bank lending." The Quarterly Journal of Economics 127.2: 829-887. Lane, Philip R. 2003. "Business Cycles and Macroeconomic Policy in Emerging Market Economies." International Finance 89-108. Lloyd, Simon, Ed Manuel, and Konstantin Panchev. 2024. "Foreign Vulnerabilities, Domestic Risks: The Global Drivers of GDP-at-Risk." IMF Economic Review, Palgrave Macmillan;International Monetary Fund vol. 72(1), pages 335-392. Mendoza, Enrique G., and P. Marcelo Oviedo. 2006. "Fiscal Policy and Macroeconomic Uncertainty in Developing Countries: The Tale of the Tormented Insurer." NATIONAL BUREAU OF ECONOMIC RESEARCH. Miningou, Elise Wendlassida. 2019. Quality Education and the Efficiency of Public Expenditure: A Cross-Country Comparative Analysis. Policy Research Working Paper;No. 9077, Washington, DC: World Bank. Ostrom, Elinor. 2015. Governing the Commons - The Evolution of Institutions for Collective Action. Cambridge: Cambridge University Press. Riascos, Alvaro, and Carlos A. Vegh. 2003. "Procyclical Government Spending in Developing Countries: The Role of Capital Market Imperfections." Talvi, Ernesto, and Carlos A. Végh. 2005. "Tax base variability and procyclical fiscal policy in developing countries." Journal of Development Economics 156-190. Tornall, Aaron, and Philip R. Lane. 1999. "The Voracity Effect." American Economic Review 22- 46. Vegh Gramont, Carlos Alberto, Guillermo Javier Vuletin, Daniel Riera-Crichton, Juan Pablo Medina, Diego Friedheim, Luis Francisco Morano Germani, and Lucila Venturi Grosso. 2018. From Known Unknowns to Black Swans : How to Manage Risk in Latin America and the Caribbean (English). Washington, D.C.: World Bank Group. Velasco, Andrés. 1999. "2. A Model of Endogenous Fiscal Deficits and Delayed Fiscal Reforms." In Fiscal Institutions and Fiscal Performance, by edited by James M. Poterba and Jurgen von Hagen, 37-58. Chicago: University of Chicago Press. Wagner, Adolph. 1958. "Three extracts on public finance." In Classics in the theory of public finance, by Richard A. Musgrave and Alan T. Peacock, 1-15. London: Palgrave Macmillan UK. World Bank. 2013. World Development Report 2014: Risk and Opportunity—Managing Risk for Development. Washington, DC: World Bank. http://hdl.handle.net/10986/16092.