Large fiscal episodes and sustainable development some international evidence

Aizenman, Joshua; Jinjarak, Yothin; Hien Nguyen; Park, Donghyun Working Paper Large fiscal episodes and sustainable development some international evidence ADB Economics Working Paper Series, No. 644 Provided in Cooperation with: Asian Development Bank (ADB), Manila Suggested Citation: Aizenman, Joshua; Jinjarak, Yothin; Hien Nguyen; Park, Donghyun (2021) : Large fiscal episodes and sustainable development some international evidence, ADB Economics Working Paper Series, No. 644, Asian Development Bank (ADB), Manila, https://doi.org/10.22617/WPS210523-2 This Version is available at: https://hdl.handle.net/10419/259482 Standard-Nutzungsbedingungen: Die Dokumente auf EconStor dürfen zu eigenen wissenschaftlichen Zwecken und zum Privatgebrauch gespeichert und kopiert werden. Sie dürfen die Dokumente nicht für öffentliche oder kommerzielle Zwecke vervielfältigen, öffentlich ausstellen, öffentlich zugänglich machen, vertreiben oder anderweitig nutzen. Sofern die Verfasser die Dokumente unter Open-Content-Lizenzen (insbesondere CC-Lizenzen) zur Verfügung gestellt haben sollten, gelten abweichend von diesen Nutzungsbedingungen die in der dort genannten Lizenz gewährten Nutzungsrechte. Terms of use: Documents in EconStor may be saved and copied for your personal and scholarly purposes. You are not to copy documents for public or commercial purposes, to exhibit the documents publicly, to make them publicly available on the internet, or to distribute or otherwise use the documents in public. If the documents have been made available under an Open Content Licence (especially Creative Commons Licences), you may exercise further usage rights as specified in the indicated licence. https://creativecommons.org/licenses/by/3.0/igo/ ASIAN DEVELOPMENT BANK ASIAN DEVELOPMENT BANK 6 ADB Avenue, Mandaluyong City 1550 Metro Manila, Philippines www.adb.org Large Fiscal Episodes and Sustainable Development Some International Evidence The link between fiscal policy and sustainable development is country specific. Tax revenues remain a major source of green and social finance. Analysis of episodes of fiscal expansion and tightening in four emerging markets finds that the association between fiscal aggregates and domestic resource mobilization and sustainable development differs across countries. Further, the time paths of tax and expenditure components are important for domestic resource mobilization to target sustainable development outcomes. More granular fiscal data may shed more light on the association between fiscal conditions and sustainable development in the coming years. About the Asian Development Bank ADB is committed to achieving a prosperous, inclusive, resilient, and sustainable Asia and the Pacific, while sustaining its efforts to eradicate extreme poverty. Established in 1966, it is owned by 68 members —49 from the region. Its main instruments for helping its developing member countries are policy dialogue, loans, equity investments, guarantees, grants, and technical assistance. LARGE FISCAL EPISODES AND SUSTAINABLE DEVELOPMENT SOME INTERNATIONAL EVIDENCE Joshua Aizenman, Yothin Jinjarak, Hien Nguyen, and Donghyun Park ADB ECONOMICS WORKING PAPER SERIES NO. 644 December 2021 ASIAN DEVELOPMENT BANK ADB Economics Working Paper Series Large Fiscal Episodes and Sustainable Development Some International Evidence Joshua Aizenman, Yothin Jinjarak, Hien Nguyen, and Donghyun Park No. 644 | December 2021 Joshua Aizenman is the Robert R. and Katheryn A. Dockson Chair in Economics and International Relations and professor of International Relations and Economics at the University of Southern California. Yothin Jinjarak is a senior economist and Donghyun Park is a principal economist at the Economic Research and Regional Cooperation Department, Asian Development Bank. Hien Nguyen is a modelling analyst at New Zealand Treasury. Creative Commons Attribution 3.0 IGO license (CC BY 3.0 IGO) © 2021 Asian Development Bank 6 ADB Avenue, Mandaluyong City, 1550 Metro Manila, Philippines Tel +63 2 8632 4444; Fax +63 2 8636 2444 www.adb.org Some rights reserved. Published in 2021. ISSN 2313-6537 (print), 2313-6545 (electronic) Publication Stock No. WPS210523-2 DOI: http://dx.doi.org/10.22617/WPS210523-2 The views expressed in this publication are those of the authors and do not necessarily reflect the views and policies ofthe Asian Development Bank (ADB) or its Board of Governors or the governments they represent. ADB does not guarantee the accuracy of the data included in this publication and accepts no responsibility for any consequence of their use. The mention of specific companies or products of manufacturers does not imply that they are endorsed or recommended by ADB in preference to others of a similar nature that are not mentioned. By making any designation of or reference to a particular territory or geographic area, or by using the term “country” inthis document, ADB does not intend to make any judgments as to the legal or other status of any territory or area. This work is available under the Creative Commons Attribution 3.0 IGO license (CC BY 3.0 IGO) https://creativecommons.org/licenses/by/3.0/igo/. By using the content of this publication, you agree to be bound bytheterms of this license. For attribution, translations, adaptations, and permissions, please read the provisions andterms of use at https://www.adb.org/terms-use#openaccess. This CC license does not apply to non-ADB copyright materials in this publication. If the material is attributed toanother source, please contact the copyright owner or publisher of that source for permission to reproduce it. ADB cannot be held liable for any claims that arise as a result of your use of the material. Please contact [email protected] if you have questions or comments with respect to content, or if you wish toobtain copyright permission for your intended use that does not fall within these terms, or for permission to use theADB logo. Corrigenda to ADB publications may be found at http://www.adb.org/publications/corrigenda. Notes: In this publication, “$” refers to United States dollars. The ADB Economics Working Paper Series presents data, information, and/or findings from ongoing research and studies to encourage exchange of ideas and to elicit comment and feedback about development issues in Asia and the Pacific. Since papers in this series are intended for quick and easy dissemination, the content may or may not be fully edited and may later be modified for final publication. CONTENTS TABLES AND FIGURES iv ABSTRACT v I. INTRODUCTION 1 II. DATA 2 III. FISCAL AGGREGATES AND SUSTAINABLE DEVELOPMENT 3 IV. LINKING FISCAL EPISODES TO SUSTAINABLE DEVELOPMENT 11 A. Identifying Fiscal Episodes 11 B. Fiscal Episodes, Gross Domestic Product Growth, and Development Outcomes 13 C. Large Fiscal Impulses and Tax-Spending Components 15 V. CONCLUSION 19 APPENDIX 20 REFERENCES 32 TABLES AND FIGURES TABLES 1 Large Fiscal Impulses and Gross Domestic Product Growth 13 2 Fiscal Stimuli, Growth, and Development Outcomes 14 3 Fiscal Adjustment, Growth, and Development Outcomes 15 4 Large Fiscal Impulses and Tax Components 16 5 Large Fiscal Impulses and Expense Components 18 6 Tax/Expense Components and Fiscal Stimuli/Adjustment 19 A1 Statistics Description of Fiscal Impulse/Stimuli/Adjustment by Country 21 A2 Fiscal Impulses and Gross Domestic Product Growth 22 A3 Fiscal Stimuli, Growth, and Development Outcomes 23 A4 Fiscal Adjustment, Growth, and Development Outcomes 24 A5 Future Fiscal Stimuli, Growth, and Development Outcomes 25 A6 Future Fiscal Adjustment, Growth, and Development Outcomes 26 A7 Past Fiscal Stimuli, Growth, and Development Outcomes 27 A8 Past Fiscal Adjustment, Growth, and Development Outcomes 28 A9 Large Fiscal Impulses and Tax Components 29 A10 Large Fiscal Impulses and Expense Components 30 A11 Tax/Expense Components and Fiscal Stimuli/Adjustment 31 FIGURES 1 Fiscal Expenses and Revenues, 2000–2018 4 2 Components of Revenues and Expenses, 2000–2018 6 3 Fiscal Components and Development Dimensions 9 4 Strong Fiscal Impulses and Gross Domestic Product Growth 12 ABSTRACT This paper examines the association between episodes of large fiscal impulses (expansions and adjustments) and sustainable development indicators (prosperity, resilience, and inclusivity). We provide country studies of Chile, Poland, South Africa, and Thailand, examining the components of government expenses and tax revenues, and reporting four stylized patterns from the analysis as follows: (i) Fiscal expansions led to higher growth rates and reduced negative trade-offs, e.g., pollution and poor-health mortalities associated with economic growth. (ii) Fiscal adjustments led to a more inclusive economy, lowered poverty headcounts, improved sanitation, and increased cleaner technology access. (iii) Fiscal expansions followed an increase in direct taxes (especially corporate taxes) and a decline in social contributions, and preceded a decline in other direct taxes and an increase in wage bills. (iv) Fiscal adjustments followed a decline in other direct taxes and social contributions, an increase in wage bills, and preceded a decline in government consumption expenditure and transfers. In light of these findings, domestic resource mobilization should consider the time paths of the taxes and expenditure components to understand their empirical linkages with sustainable development outcomes in the respective countries. Keywords: sustainable development, tax base, government expenses, large fiscal changes JEL codes: E62, H11, O11 I. INTRODUCTION This paper studies the intertemporal feedbacks between fiscal policy and sustainable development. We focus on the inclusivity of economic growth and domestic revenue mobilization around large fiscal mobilization episodes, including reforms of indirect taxes and exemptions, tax administration, and transfer programs. Inclusive growth reduces poverty, improves health and education, among others, and may enhance the tax capacity, spending efficiency, and resource mobilization (e.g., greater compliance, higher income, lower informality). In turn, higher tax mobilization may fund pressing social policies and redistribution, thereby deepening inclusive growth and better social infrastructure. Such associations are subject to structural factors, including governance, polarization, civil wars and conflicts, ethnic and religious fragmentation, quality of institutions, a share of manufacturing, farming, commodity dependence, and urbanization rates. To deepen our understanding amid the challenges posed by multidimensional data requirements, we provide country studies that are combined with a panel estimation of selected emerging markets. We ask three questions. The first is the relationship between economic structure and the composition of government revenues and expenses. Second, we address the association between episodes of large fiscal impulses (expansions and adjustments) and sustainable development outcomes (prosperity, inclusivity, and resilience). Third, we ask whether past changes in the budget’s specific components led to fiscal expansions and adjustments, which relied on future changes in certain budget items. After putting together the necessary data, we identify Chile, Poland, South Africa, and Thailand as the emerging markets that provide sufficient fiscal aggregates and represent a diverse sample set for the analysis. For these countries, a third of gross domestic product (GDP) is the industry, and more than half are services, based on the World Bank’s World Development Indicators database as of 2018. Despite the similarity, there are structural differences. Chile’s largest exports are metals, minerals, and agricultural products; South Africa’s largest exports are stone, minerals, and agricultural products based on The Growth Lab at Harvard University (2019) as of 2018: the fiscal stance is vulnerable to commodity price fluctuations. Poland’s largest exports are machinery, services, and agricultural products; Thailand’s largest exports are services, machinery, electronics, vehicles, and chemicals: the fiscal conditions are sensitive to global industrial production and related supply-chain services. Thus, different terms of trade shocks are responsible for large episodes of fiscal expansions and adjustments across these economies. There are also differences in their politico-economic structure, i.e., liberal versus conservative tendency and single-party versus coalition government. These subtleties are beyond the scope of our paper. The analysis proceeds with the decomposition of fiscal revenues and expenses and their linkages with sustainable development. The estimation then tests these linkages’ strengths, identifying prevalent useful feedback as the guidelines for dealing with the present and future challenges. We find some stylized patterns in the sample countries: (i) fiscal expansions led to higher growth rates and reduced negative trade-offs, e.g., pollution and poor-health mortalities associated with economic growth; (ii) fiscal adjustments led to a more inclusive economy, lowered poverty headcounts, improved sanitation, and cleaner technology access; (iii) fiscal expansions followed an increase in direct taxes (especially corporate taxes) and a decline in social contributions, and preceded a decline in other direct taxes and an increase in wage bills; and (iv) fiscal adjustments followed a decline in other direct taxes and social contributions, an increase in wage bills, and preceded a decline in government 2 ADB Economics Working Paper Series No. 644 consumption expenditure and transfers. In light of these findings, the domestic resource mobilization should consider the time paths of the taxes and expenditure components to understand their empirical linkages with the sustainable development outcomes in the respective countries. Section II describes the data. Section III discusses fiscal aggregates and sustainable development and follows with the estimation of the linkages in section IV. The conclusion is in section V. II. DATA Data on the main components of the general government spending are from Government Finance Statistics database of the International Monetary Fund, including wage bills (Wage), nonwage expenditure (Nonwage), subsidies to firms (Subsidies), and expenditure on social benefits (Transfers); all in percent of gross domestic product (GDP). Because data on fixed capital consumption are not available for Thailand and only available for South Africa from 2003, we exclude it from the study. This exclusion should not influence our analysis as the fixed capital consumption is only a minor part of the general government expenditure (0.93% in Chile 2000–2018, 2.45% in Poland 1995–2018, and 0.7% in South Africa 2003–2018). Following the literature, we also exclude interest expense to focus on the discretionary change of fiscal impulse. Our data indicate that the interest expense accounts for a small fraction of the general government expenses (0.76% in Chile 2000–2018, 2.78% in Poland 1995–2018, 3.81% in South Africa 1996–2018, and 1.1% in Thailand 2000–2018). We use the Government Revenue Dataset of the United Nations University World Institute for Development Economics Research (UNU-WIDER) for the main components of total tax revenue. The data include personal income taxes (PIT), corporate income taxes (CIT), payroll and workforce taxes and property taxes (Other), indirect taxes (INDT), and social security contributions (SSC); all in percent of GDP. For Chile, we interpolate the 1993–2005 nonresource CIT from the total CIT—total income and profit taxes on corporations, including taxes on resource firms—over the 1993–2005 period, using the 2006–2018 ratio of the nonresource CIT to total CIT.1 We calculate the primary balance as follows: •Total tax revenue (% of GDP) = PIT + CIT + Other + INDT + SSC •Primary expenses (% of GDP) = Wage + Nonwage + Subsidies + Transfer •Primary deficit (% of GDP) = Primary expenses - Total tax revenue Our analysis considers economic growth, taking into account its inclusivity and sustainability. For comprehensiveness, we use 17 Sustainable Development Goals (SDGs) indicators, covering poverty headcount ratio (Poverty); population share of undernourished (Undernourishment); mortality ratio from cardiovascular disease (CVD), cancer, diabetes, or chronic respiratory disease (CRD) between age 30 and age 70 (Mortality); enrollment ratio to pre-primary school (Enrollment); proportion of seats held by women in national parliaments (Women power); population share with access to basic sanitation services (Sanitation); population share with access to clean fuels and 1 Specifically, 𝐶𝐼𝑇, = ∑, ,   ×𝑇𝐶𝐼𝑇 , , x where t = 1993–2005; TCIT is total income and profit taxes on corporations, including taxes on resource firms. Large Fiscal Episodes and Sustainable Development 9 Figure 3: Fiscal Components and Development Dimensions (Sustainable Development Goals) continued on next page –1.0 –0.5 0.0 0.5 1.0 No value Pearson Correlation *** *** ** *** *** *** ** ** *** *** *** ** ** *** ** *** *** ** *** ** *** *** *** *** *** *** *** ** *** ** *** *** Wage Nonwage Subsidies Transfer PIT CIT Other INDT SSC Primary expenses Tax revenue Primary deficit Wage Nonwage Subsidies Transfer PIT CIT Other INDT SSC Primary expenses Tax revenue Primary deficit Expenses and Revenue components *** ** ** ** ** *** *** ** *** ** ** ** ** *** * *** *** *** *** *** *** *** *** ** ** ** ** ** *** *** ** ** ** *** *** ** ** ** *** ** *** * *** *** ** ** *** ** *** *** *** *** *** Poverty Undernourishment Mortality Enrollment Women power Sanitation Fuel-tech Non-education Broadband Bottom consumption Pollution Carbon dioxide damage Displacement Marine Terrestrial Bribery Remittances Wage Nonwage Subsidies Transfer PIT CIT Other INDT SSC SDGs and Expenses/Revenue components a. Chile 2000–2018 *** *** *** *** *** *** *** *** *** *** *** *** *** *** *** *** *** *** *** *** *** *** *** *** *** *** *** *** *** Wage Nonwage Subsidies Transfer PIT CIT Other INDT SSC Primary expenses Tax revenue Primary deficit Wage Nonwage Subsidies Transfer PIT CIT Other INDT SSC Primary expenses Tax revenue Primary deficit Expenses and Revenue components *** *** *** *** *** *** *** *** ** *** *** *** *** *** *** *** *** *** *** *** ** *** *** ** *** *** ** *** *** *** *** *** ** *** *** ** *** * *** *** *** ** *** *** *** *** *** *** *** ** *** *** *** ** ** ** *** *** *** Poverty Undernourishment Mortality Enrollment Women power Sanitation Fuel−tech Non-education Broadband Bottom consumption Pollution Carbon dioxide damage Displacement Marine Terrestrial Bribery Remittances SDGs and Expenses/Revenue components b. Poland 1995–2018 Wage Nonwage Subsidies Transfer PIT CIT Other INDT SSC 10 ADB Economics Working Paper Series No. 644 Table 3 continued CIT = nonresource corporate income taxes, INDT = indirect taxes, Nonwage = expense on good and services, Other = taxes on payroll and workforce and taxes on the property, PIT = nonresource personal income taxes, SDG = Sustainable Development Goal, SSC = social security contributions, Subsidies = subsidies to firms, Transfer = expense on social benefits, Wage = compensation on employees. Notes: 1. Statistical associations greater than 0.5 have asterisk(s) in the heatmaps; *** = p<0.01, ** = p<0.05, * = p<0.1. 2. The associations between fiscal expenses, taxes, and sustainable development outcomes differed across Chile, Poland, South Africa, and Thailand. The differences motivate a focal study on fiscal episodes as the lowest-hanging fruit in understanding the empirical linkages between fiscal aggregates and domestic resource mobilization’s role in inclusive growth and sustainable development. Sources: International Monetary Fund. Government Finance Statistics. https://data.imf.org/?sk=A0867067-D23C-4EBCAD23-D3B015045405 (accessed 7 September 2020); UNU-WIDER. Government Revenue Dataset. https://doi.org/10.35188/UNU-WIDER/GRD-2021 (accessed 7 September 2020). *** *** *** *** *** *** *** *** *** *** *** *** *** *** *** *** *** *** *** ** *** *** *** *** *** *** *** *** Wage Nonwage Subsidies Transfer PIT CIT Other INDT SSC Primary expenses Tax revenue Primary deficit Wage Nonwage Subsidies Transfer PIT CIT Other INDT SSC Primary expenses Tax revenue Primary deficit Expenses and Revenue components *** *** *** ** *** *** *** *** *** *** *** *** *** ** *** *** ** ** ** *** *** *** *** ** *** *** *** *** *** *** *** ** *** *** *** ** *** *** *** *** *** *** *** *** *** *** *** *** ** *** *** *** *** ** *** Poverty Undernourishment Mortality Enrollment Women power Sanitation Fuel-tech Non-education Broadband Pollution Carbon dioxide damage Displacement Marine Terrestrial Remittances Wage Nonwage Subsidies Transfer PIT CIT Other INDT SSC SDGs and Expenses/Revenue components c. South Africa 1996−2018 *** *** *** ** *** *** *** *** *** *** *** ** *** ** *** *** *** ** ** *** *** *** *** ** *** Wage Nonwage Subsidies Transfer PIT CIT Other INDT SSC Primary expenses Tax revenue Primary deficit Wage Nonwage Subsidies Transfer PIT CIT Other INDT SSC Primary expenses Tax revenue Primary deficit Expenses and Revenue components ** ** ** ** *** *** *** *** *** ** *** *** *** *** ** ** ** ** ** ** ** ** ** *** *** ** *** ** *** *** ** ** ** * *** *** *** ** ** *** * *** ** * * *** *** *** *** *** ** *** *** Poverty Undernourishment Mortality Enrollment Women power Sanitation Fuel-tech Non-education Broadband Bottom consumption Pollution Carbon dioxide damage Displacement Marine Terrestrial Remittances Wage Nonwage Subsidies Transfer PIT CIT Other INDT SSC SDGs and Expenses/Revenue components d. Thailand 2000−2018 –1.0 –0.5 0.0 0.5 1.0 No value Pearson Correlation Large Fiscal Episodes and Sustainable Development 11 IV. LINKING FISCAL EPISODES TO SUSTAINABLE DEVELOPMENT A. Identifying Fiscal Episodes Following Alesina and Ardagna (2010) and Alesina et al. (1995), we define fiscal episodes using the significant change in the primary deficit (as a share of GDP) from the previous year. The underlying assumption is that the unemployment rate remained stable in period t as in period t-1. According to Blanchard (1990), this approach takes the previous year as the benchmark period (t-1) and estimates the government expenses and tax revenue in the current year (t), thereby filtering out the variations in fiscal variables induced by business cycle fluctuations. Subject to data availability, we follow this approach for simplicity, avoiding the challenges posed by country-specific calculation of potential outputs. To define the cyclically adjusted variables and episodes of large fiscal impulse: • Cyclically adjusted Primary expenses (% of GDP) = Cyclically adjusted Transfer + (Wage + Nonwage + Subsidies) • Cyclically adjusted Total tax revenue (% of GDP) = Cyclically adjusted PIT + Cyclically adjusted CIT + Other + Cyclically adjusted INDT + Cyclically adjusted SSC • Cyclically adjusted Primary deficit (% of GDP) = Cyclically adjusted Primary expenses – Cyclically adjusted Total tax revenue • Fiscal impulset (% of GDP) = Cyclically adjusted Primary deficitt – Primary deficitt-1 We correct each of the fiscal variables at period t (denoted as 𝑋) the variations caused by cyclical factors. For each country in the sample, we first regress each fiscal variable (𝑋) on a time trend (𝑇𝑅𝐸𝑁𝐷) and the unemployment rate (𝑈). 𝑋=𝛼+𝛼𝑇𝑅𝐸𝑁𝐷+𝛼𝑈+𝜀 (1) Next, we estimate what each fiscal variable would be in period t if the unemployment rate were to remain the same as in the previous year (𝑈), i.e., the cyclically adjusted fiscal variable. 𝑋(𝑎𝑡 𝑈)=𝛼 +𝛼  𝑇𝑅𝐸𝑁𝐷+𝛼 𝑈 +𝜀  (2) where the terms, 𝛼 , 𝛼 , 𝛼 , and 𝜀  are coefficient estimates. Having estimated 𝑇𝑟𝑎𝑛𝑠𝑓𝑒𝑟(𝑎𝑡 𝑈), 𝑃𝐼𝑇(𝑎𝑡 𝑈), 𝐶𝐼𝑇(𝑎𝑡 𝑈), 𝐼𝑁𝐷𝑇(𝑎𝑡 𝑈), and 𝑆𝑆𝐶(𝑎𝑡 𝑈), we calculate 𝑃𝑟𝑖𝑚𝑎𝑟𝑦 𝑒𝑥𝑝𝑒𝑛𝑠𝑒𝑠(𝑎𝑡 𝑈), 𝑇𝑜𝑡𝑎𝑙 𝑡𝑎𝑥 𝑟𝑒𝑣𝑒𝑛𝑢𝑒(𝑎𝑡 𝑈), and then 𝑃𝑟𝑖𝑚𝑎𝑟𝑦 𝑑𝑒𝑓𝑖𝑐𝑖𝑡(𝑎𝑡 𝑈). The Fiscal impulse at period t is the difference between the cyclically adjusted primary deficit at period t [𝑃𝑟𝑖𝑚𝑎𝑟𝑦 𝑑𝑒𝑓𝑖𝑐𝑖𝑡(𝑎𝑡 𝑈)] and the actual primary deficit at period t-1 [𝑃𝑟𝑖𝑚𝑎𝑟𝑦 𝑑𝑒𝑓𝑖𝑐𝑖𝑡]. We estimate fiscal impulse using the balanced data of expenses and tax revenue components in each country: Chile 2000–2018, Poland 1995–2018, South Africa 1996–2018, and Thailand 2000–2018. We have 81 observations of fiscal impulse in the four sample countries; the mean fiscal impulse is 0.036% of GDP and a standard deviation of 1.308% of GDP. Appendix Table A1 provides 12 ADB Economics Working Paper Series No. 644 more detailed statistics. We identify episodes of large fiscal impulse (discretionary fiscal policy): fiscal adjustment and fiscal stimuli as follows: • strong fiscal adjustment if Fiscal impulse (% of GDP) is less than –1.5 • strong fiscal stimuli if Fiscal impulse (% of GDP) is larger than 1.5 Figure 4 plots episodes of large fiscal impulse and GDP growth by country. The identification results in six episodes of fiscal adjustment and six episodes of fiscal stimuli. The longest episode persisted for 2 years (fiscal stimuli in Poland during 2008–2009 and fiscal adjustment in South Africa during 2005–2006). Many fiscal stimuli episodes occurred during the 2007–2010 global financial crisis. The time path of GDP growth suggests some association with the fiscal episodes of stimuli (positive) and adjustment (negative) in the sample period. Figure 4: Strong Fiscal Impulses and Gross Domestic Product Growth GDP = gross domestic product. Note: Using the cyclically adjusted fiscal aggregates in the analysis, we identify 12 episodes of large fiscal impulses (expansions and adjustments), many of which occurred during the global financial crisis in the 1995–2018 sample period. Sources: International Monetary Fund. Government Finance Statistics. https://data.imf.org/?sk=A0867067-D23C-4EBC-AD23D3B015045405 (accessed 7 September 2020); UNU-WIDER. Government Revenue Dataset. https://doi.org/10.35188/UNUWIDER/GRD-2021 (accessed 7 September 2020). 0.0 2.5 5.0 7.5 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 GDP growth rate (%) Chile 2 4 6 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 GDP growth rate (%) Poland Episode Fiscal adjustment Fiscal stimuli 0 2 4 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 GDP growth rate (%) South Africa 0 2 4 6 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 GDP growth rate (%) Thailand Large Fiscal Episodes and Sustainable Development 13 B. Fiscal Episodes, Gross Domestic Product Growth, and Development Outcomes For our baseline estimation, we conduct fixed effects regressions on the four sample countries. First, we examine the association between GDP growth and large fiscal stimuli/adjustment changes across the years using the following equations: 𝐺𝐷𝑃 𝑔𝑟𝑜𝑤𝑡ℎ =𝜑+𝜑𝐹𝑖𝑠𝑐𝑎𝑙 𝐸𝑝𝑖𝑠𝑜𝑑𝑒 +𝑢 (3) 𝐹𝑖𝑠𝑐𝑎𝑙 𝐸𝑝𝑖𝑠𝑜𝑑𝑒 =𝛿 +𝛿𝐺𝐷𝑃 𝑔𝑟𝑜𝑤𝑡ℎ +𝜇 (4) 𝐹𝑖𝑠𝑐𝑎𝑙 𝐸𝑝𝑖𝑠𝑜𝑑𝑒 =𝛾 +𝛾𝐺𝐷𝑃 𝑔𝑟𝑜𝑤𝑡ℎ +𝑣 (5) Table 1 reports the estimates of equation 3 in column 1 and equation 4 in column 2, focusing here on the statistically significant coefficients.2 We find a positive association between current GDP growth and lagged fiscal stimuli (column 1) and between current fiscal stimuli and lead GDP growth (column 2). While these findings do not pin down causality, they suggest that GDP growth tends to follow fiscal stimuli episodes. There is no association between lagged GDP growth and fiscal adjustment (equation 5). Table 1: Large Fiscal Impulses and Gross Domestic Product Growth (1 ) (2 ) GDP GrowthtFiscal Stimulit Fiscal stimulit-1 0.398* ( 0.129) GDP growtht+1 0.009* ( 0.003) Observations 81 81 p-value 0.054 0.058 R-squared (within) 0.004 0.004 GDP = gross domestic product. Notes: 1. Fixed-effects estimation with country-fixed effects included. Robust standard errors in parentheses. * = p<0.1. 2. GDP growth followed fiscal expansions. There was no association between growth rates and fiscal adjustments. Source: Authors’ calculations. Next, we look at the associations of development outcomes, GDP growth, and the interaction between GDP growth and fiscal impulses (i.e., episodes of fiscal stimuli and adjustment). 𝑆𝐷𝐺 =𝛽 +𝛽 𝐺𝐷𝑃 𝑔𝑟𝑜𝑤𝑡ℎ +𝛽𝐹𝑖𝑠𝑐𝑎𝑙 𝐸𝑝𝑖𝑠𝑜𝑑𝑒 + (6) 𝛽(𝐺𝐷𝑃 𝑔𝑟𝑜𝑤𝑡ℎ × 𝐹𝑖𝑠𝑐𝑎𝑙 𝐸𝑝𝑖𝑠𝑜𝑑𝑒)+𝜗 2 Full estimations are in Appendix Table A2. 14 ADB Economics Working Paper Series No. 644 We regress equation 6 using both the fiscal stimuli and the fiscal adjustment, one at a time.3 The total effect of large fiscal impulses on SDGs are thus (𝛽 + 𝛽 𝐺𝐷𝑃 𝑔𝑟𝑜𝑤𝑡ℎ). Similarly, the total effect of GDP growth on SDGs are (𝛽 + 𝛽 𝐹𝑖𝑠𝑐𝑎𝑙 𝐸𝑝𝑖𝑠𝑜𝑑𝑒); where 𝛽 , 𝛽 , and 𝛽  are estimated coefficients. Having estimated equation 6 on all 17 SDG indicators, Tables 2 and 3 report the results, focusing on the development outcomes with statistically significant estimates.4 From Table 2, the current GDP growth is positively associated with current poverty headcount ratio, mortality ratio (CVD, cancer, diabetes, CRD for the people aged 30–70), and PM2.5 air pollution. This finding implies the trade-off between GDP growth and these sustainable development goals in the country sample. Current fiscal stimuli episodes are positively associated with the current poverty headcount ratio and PM2.5 air pollution and are negatively associated with the number of newly displaced people related to disasters. This finding suggests the impacts of fiscal stimuli on sustainable development goals differ between the long-term/persistent targets (poverty, health, pollution) and the short-term/pressing issues (disasters). Higher growth associated with the fiscal stimuli is negatively associated with the mortality rate from poor health conditions. Given the association between current GDP growth and lagged fiscal stimuli (Table 1), this finding implies that fiscal stimuli help mitigate the trade-off between GDP growth and some SDG outcomes. Table 2: Fiscal Stimuli, Growth, and Development Outcomes (1) (2 ) (3 ) (4 ) Povertyt MortalitytPollutiontDisplacementt GDP growtht 1.822* 0.239* 0.387* 0.009 ( 0.589) ( 0.077) ( 0.137) (0.024) Fiscal stimulit 6.786** 0.749 2.477** –0.247* ( 1.910) ( 0.386) ( 0.601) (0.091) Fiscal stimulit –2.055 –0.248* –0.297 –0.045 x GDP growth ratet ( 1.114) ( 0.084) ( 0.228) (0.024) Observations 60 66 77 42 p-value 0.042 0.019 0.053 0.000 R-squared (within) 0.189 0.100 0.202 0.040 GDP = gross domestic product. Notes: 1. Fixed-effects estimation with country-fixed effects included. Robust standard errors in parentheses. ** = p<0.05, * = p<0.1. 2. The linkages of fiscal expansions and sustainable development outcomes differed between the long-term objectives/persistent issues (i.e., poverty, health, pollution) and the short-term objectives/urgent issues (i.e., disasters). Fiscal expansions reduced the negative trade-off between GDP growth and some development indicators, e.g., pollution and poor-health mortalities associating with economic growth. Source: Authors’ calculations. From Table 3, the GDP growth rate is positively associated with the poverty headcount ratio and mortality from CVD, cancer, diabetes, or CRD for the people aged 30–70. However, GDP growth is negatively correlated with broadband subscriptions. Fiscal adjustment episodes are negatively 3 Both the fiscal stimuli or adjustment and economic growth are considered endogenous regressors. As shown in Table 1, the lags and leads of these variables are correlated. The two stage-least-squares estimation can be appropriately done in a larger than the current sample, with more economies and time periods. 4 Full estimations for 17 SDGs are in Appendix Tables A3–A4. Large Fiscal Episodes and Sustainable Development 15 associated with the poverty headcount ratio and are positively associated with access to basic sanitation services and clean fuels and cooking technologies. The association between fiscal adjustment episodes and poverty headcount ratio declines as GDP grows: the threshold of GDP growth is 5.24% (12.014 ÷ 2.294), above which the effect of fiscal adjustment dissipated. The GDP growth coinciding with fiscal adjustment episodes is negatively associated with access to clean fuels and technologies for cooking (the GDP growth threshold is 5.96%). It is positively associated with the young population share, not in education, employment, or training. Table 3: Fiscal Adjustment, Growth, and Development Outcomes (1) (2 ) (3 ) (4 ) (5) (6 ) Poverty Mortality Sanitation Fuel-Tech Non-Education Broadband GDP growth ratet 1.071** 0.194** –0.558 –0.677 –0.093 –0.462* ( 0.257) ( 0.044) ( 0.309) ( 0.509) (0.182) ( 0.150) Fiscal adjustmentt –12.014** –0.819 3.261** 9.805*** –3.060 5.485 ( 3.055) ( 0.506) ( 0.933) ( 1.591) ( 1.500) ( 2.754) Fiscal adjustmentt 2.294*** 0.054 –0.465 –1.644** 0.640* –0.641 x GDP growth ratet ( 0.211) ( 0.056) ( 0.248) ( 0.460) ( 0.230) (0.361) Observations 60 66 70 66 62 71 p-value 0.000 0.000 0.013 0.000 0.000 0.011 R-squared (within) 0.170 0.089 0.118 0.115 0.008 0.056 GDP = gross domestic product. Notes: 1. Fixed-effects estimation with country-fixed effects included. Robust standard errors in parentheses. *** = p<0.01, ** = p<0.05, * = p<0.1. 2. Lowered poverty headcounts and improved sanitation, and cleaner technology followed fiscal adjustments. However, there was some trade-off between GDP growth and such association in the sample. Source: Authors’ calculations. We also adjust equation 6 by entering lead (period t+1) and lagged (period t-1) variables of GDP growth, large fiscal impulses, and the interaction between GDP growth and large fiscal impulses as regressors. Appendix Tables A5 and A6 report detailed results using lead explanatory variables and Tables A7 and A8 for lagged variables. The extended results are largely supportive to the baseline findings from equation 6 regression for the four sample countries that: (i) there is the trade-off between (lagged, current, and lead) GDP growth rate and the development goals; (ii) the fiscal adjustment (stimuli) episodes are favorable (or unfavorable) to development goals; and (iii) the GDP growth coinciding with fiscal adjustment (stimuli) episodes deteriorates (or improves) the development goals. These findings are not applicable for all SDGs, however. C. Large Fiscal Impulses and Tax-Spending Components We further test whether fiscal episodes are associated with any components of the fiscal trends by estimating the following equation of fiscal episode at time t on the fiscal components at time t+1 and t-1: 𝐹𝑖𝑠𝑐𝑎𝑙 𝑒𝑝𝑖𝑠𝑜𝑑𝑒 =𝜇+𝜇𝑇𝑎𝑥 (𝑜𝑟 𝐸𝑥𝑝𝑒𝑛𝑠𝑒)𝑐𝑜𝑚𝑝𝑜𝑛𝑒𝑛𝑡𝑠 +𝜀 (7) 𝐹𝑖𝑠𝑐𝑎𝑙 𝑒𝑝𝑖𝑠𝑜𝑑𝑒 =𝜃 +𝜃𝑇𝑎𝑥 (𝑜𝑟 𝐸𝑥𝑝𝑒𝑛𝑠𝑒)𝑐𝑜𝑚𝑝𝑜𝑛𝑒𝑛𝑡𝑠 +𝜖 (8) 16 ADB Economics Working Paper Series No. 644 We estimate equations 7 and 8 separately for fiscal stimuli and fiscal adjustment episodes. The independent variables include cyclically adjusted tax (or expense) components as a share of GDP. We report the significant findings on the tax components’ estimates in Table 4 and the expense elements in Table 5.5 Note that entering each (lead and lagged) component as an explanatory variable at a time does not alter the significant findings below. The results are available upon request. Table 4: Large Fiscal Impulses and Tax Components Fiscal StimulitFiscal Adjustmentt (1) (2 ) (3 ) (4 ) (5) (6 ) Personal income taxt+1 0.058 ( 0.049) Corporate income taxt+1 0.003 ( 0.056) Other direct taxest+1 –0.242 ( 0.220) Direct taxest+1 0.025 ( 0.050) Indirect taxest+1 –0.077 –0.072 ( 0.055) ( 0.049) Social contributionst+1 –0.218** –0.232** (0.066) ( 0.060) Personal income taxt–1 0.019 –0.054 (0.037) (0.030) Corporate income tax t–1 0.095*** –0.039 (0.007) (0.036) Other direct taxes t–1 –0.225*** –0.632*** (0.027) (0.086) Direct taxes t–1 0.067** –0.043 (0.018) (0.045) Indirect taxes t–1 0.031 0.015 –0.007 –0.014 (0.092) (0.076) (0.060) (0.064) Social contributions t–1 –0.089** –0.087** –0.103* –0.135** (0.020) (0.017) (0.037) (0.035) Observations 77 77 81 81 81 81 p-value … 0.010 … 0.000 … 0.099 R-squared (within) 0.142 0.124 0.137 0.097 0.124 0.059 … = no value. Notes: 1. Explanatory variables are cyclically adjusted. Fixed effects estimation with country-fixed effects included. Robust standard errors in parentheses. *** = p<0.01, ** = p<0.05, * = p<0.1. 2. The time paths of fiscal aggregates differ across countries. Fiscal expansions followed an increase in direct taxes, especially corporate tax, and a decline in social contributions. Fiscal adjustments followed a decline in other direct taxes and social contributions. Source: Authors’ calculations. 5 Full estimations are in Appendix Tables A9–A10. Large Fiscal Episodes and Sustainable Development 17 We note that tax components, including personal income tax, corporate income tax, indirect taxes, and social contributions, are cyclically adjusted but not other direct taxes; all are in GDP percentage. However, a proper analysis requires more details on the budgetary process to account for the decisions on each of the items and square them with the stated fiscal stance. Fiscal stance is typically determined at the beginning of a fiscal year as part of the government’s budget cycle based on macroeconomic and revenue projections, and the fiscal adjustments in the middle of the fiscal year may not be consistent with the budget credibility underlying the fiscal management practices. Columns 1–4 in Table 4 show that fiscal stimuli episodes (period t) are negatively associated with lead and lagged social contributions (periods t+1, t-1), suggesting that the episodes of fiscal stimuli are associated with trends in the social contributions. Fiscal stimuli episodes are positively associated with lagged corporate income taxes and negatively associated with lagged other taxes, both of which were not subject to these tax components’ trend. As an additional test, we aggregate cyclically adjusted direct taxes and re-estimate equations 7–8. Specifically, Cyclically adjusted direct tax (% of GDP) = Cyclically adjusted PIT + Cyclically adjusted CIT + Other. Columns 2 and 4 of Table 4 provide the results, further supporting the negative association between fiscal stimuli and social contributions. Fiscal stimuli episodes are also positively associated with lagged direct taxes. For the fiscal adjustments, we find that they are negatively associated with lagged other direct taxes (column 5 of Table 4) and lagged social contributions (column 6 of Table 4). Fiscal adjustments are not associated with any lead tax components. Table 5 reports the significant estimates of equations 7 and 8 for fiscal adjustment episodes and expense components. Fiscal stimuli episodes are not associated with any (lead and lagged) expenditure components. We include four expense components, including wage bills, nonwage bills, subsidies expenditure, and cyclically adjusted transfer (all as a share of GDP). Columns 1 and 3 of Table 5 show that fiscal adjustment episodes (period t) are negatively associated with lead transfer expenses (period t+1) but not with the lagged transfer expense (period t-1). Aggregating government consumption expenditure: Government consumption expenditure (% of GDP) = Wage + Nonwage, shown in column 2, we find that the lead transfer component (period t+1) is still negatively associated with current fiscal adjustment episodes (period t). Besides, lead government consumption expenditure is negatively associated with current fiscal adjustment episodes. We also find that current fiscal adjustment episodes (period t) are positively associated with the lagged expense on wage bills (period t-1), shown in column 3. We also test whether current tax or expense components (period t) are associated with lagged fiscal episodes (period t-1) by estimating the following equation: 𝑇𝑎𝑥 (𝑜𝑟 𝐸𝑥𝑝𝑒𝑛𝑠𝑒)𝑐𝑜𝑚𝑝𝑜𝑛𝑒𝑛𝑡𝑠 =𝜋+𝜋𝐹𝑖𝑠𝑐𝑎𝑙 𝐸𝑝𝑖𝑠𝑜𝑑𝑒 +𝜖 (9) 18 ADB Economics Working Paper Series No. 644 Table 5: Large Fiscal Impulses and Expense Components Fiscal Adjustmentt (1 ) (2 ) (3 ) Waget+1 0.020 ( 0.031) Nonwaget+1 –0.059 ( 0.056) Government consumption expendituret+1 –0.016* ( 0.006) Subsidiest+1 –0.070 –0.083 ( 0.060) ( 0.048) Transfert+1 –0.060* –0.073** ( 0.022) ( 0.015) Waget–1 0.042* (0.016) Nonwage t–1 0.003 (0.017) Subsidies t–1 –0.009 (0.025) Transfer t–1 –0.071 (0.042) Observations 77 77 81 p-value ... 0.061 ... R-squared (within) 0.084 0.074 0.029 … = no value. Notes: 1. Explanatory variables are cyclically adjusted. Fixed effects estimation with country-fixed effects included. Robust standard errors in parentheses. ** = p<0.05, * = p<0.1. 2. Fiscal adjustments followed an increase in wage bills and preceded a decline in government consumption expenditure and transfers. Source: Authors’ calculations. As shown in Table 6 column 1, the other direct taxes (period t) are negatively associated with lagged fiscal stimuli (period t-1); in contrast, wage bills are positively associated with the lagged fiscal stimuli (column 2). We also find that cyclically adjusted transfer expense is negatively associated with the lagged fiscal adjustment (column 3). Appendix Table A11 provides full estimation results across tax and expense components. Appendix 25 Table A5: Future Fiscal Stimuli, Growth, and Development Outcomes (1) (2 ) (3 ) (4 ) (5 ) (6 ) (7 ) (8 ) (9 ) Povertyt UndernourishmenttMortalitytEnrollmenttWomen Powert SanitationtFuel-TechtNon-EducationtBroadbandt GDP growtht+1 1.953* 0.237 0.217 –0.160 –0.467 –0.730 –0.914 –0.223 –0.901** (0.626) (0.252) (0.100) (0.598) (0.331) (0.498) (0.712) (0.668) (0.207) Fiscal stimulit+1 8.796** 0.570 0.827 0.700 –2.073 –2.995 –2.971 –0.931 –3.752*** (2.307) (0.749) (0.452) (3.779) (2.076) (1.731) (2.324) (2.505) (0.611) Fiscal stimulit+1 –2.379* –0.188 –0.199 0.103 0.372 0.689 0.795 –0.290 0.854 x GDP growth ratet+1 (0.950) (0.214) (0.098) (1.053) (0.517) (0.499) (0.608) (0.622) (0.454) Observations 60 72 68 81 78 72 68 59 68 p-value 0.013 0.718 0.000 0.004 0.151 0.036 0.670 0.001 0.001 R-squared (within) 0.183 0.046 0.080 0.002 0.032 0.124 0.118 0.018 0.092 (10) (11 ) (12 ) (13 ) (14) (15 ) (16 ) (17) Bottom Consumptiont Pollutiont Carbon Dioxide DamagetDisplacementtMarinetTerrestrialtBriberytRemittances GDP growtht+1 0.681* 0.235 –0.003 0.053** 5.468*** 0.503 –0.428 0.008 (0.219) (0.132) (0.037) ( 0.015) (0.591) (0.778) (0.515) ( 0.017) Fiscal stimulit+1 0.000 1.832** –0.203* –0.094 0.000 0.000 –2.853 0.048 … (0.553) (0.084) ( 0.101) … … (3.432) ( 0.318) Fiscal stimulit+1 0.000 –0.101 0.033 –0.095** 0.000 0.000 0.000 –0.021 x GDP growth ratet+1 … (0.199) (0.025) ( 0.018) … … … ( 0.020) Observations 15 81 81 38 8 8 16 81 p-value 0.090 0.017 0.034 0.000 0.003 0.564 ... 0.131 R-squared (within) 0.614 0.109 0.011 0.078 0.931 0.055 0.054 0.003 … = no value; GDP = gross domestic product. Note: Fixed-effects estimation with country-fixed effects included. Robust standard errors in parentheses. *** = p<0.01, ** = p<0.05, * = p<0.1. Source: Authors’ calculations. 26 Appendix Table A6: Future Fiscal Adjustment, Growth, and Development Outcomes (1) (2 ) (3 ) (4 ) (5 ) (6 ) (7 ) (8 ) (9 ) Povertyt UndernourishmenttMortalitytEnrollmenttWomen Powert SanitationtFuel-TechtNon-EducationtBroadbandt GDP growtht+1 1.076** 0.215 0.165* –0.432 –0.410 –0.489 –0.649 –0.292 –0.652** ( 0.296) ( 0.216) ( 0.063) ( 0.537) ( 0.252) ( 0.328) ( 0.477) ( 0.578) ( 0.141) Fiscal adjustmentt+1 –12.421*** –0.422 –0.671 –1.011 –3.994 3.160** 3.565 –3.535 4.978 ( 1.754) ( 0.658) ( 0.430) ( 3.308) ( 2.980) ( 0.884) ( 2.657) ( 2.186) ( 2.766) Fiscal adjustmentt+1 2.330*** –0.045 0.058 1.401 0.754 –0.602* –0.691 0.822 –0.624 x GDP growth ratet+1 ( 0.252) ( 0.216) ( 0.055) ( 1.108) ( 0.529) ( 0.247) ( 0.696) ( 0.555) ( 0.452) Observations 60 72 68 81 78 72 68 59 68 p-value 0.000 0.002 0.006 0.001 0.003 0.001 0.018 0.000 0.027 R-squared (within) 0.142 0.049 0.067 0.031 0.039 0.098 0.102 0.021 0.084 (10) (11 ) (12 ) (13 ) (14) (15 ) (16 ) (17) Bottom Consumptiont Pollutiont Carbon Dioxide DamagetDisplacementtMarinetTerrestrialtBriberytRemittancest GDP growtht+1 0.777** 0.118 0.016 0.014 0.024 6.771 –0.079 0.001 ( 0.119) ( 0.053) ( 0.036) (0.033) ( 0.015) ( 3.926) ( 0.175) ( 0.002) Fiscal adjustmentt+1 1.499*** –1.392 –0.650 –0.735** –16.153*** 18.598 35.943*** –0.081 ( 0.148) ( 1.679) ( 0.426) (0.164) ( 0.042) ( 10.839) ( 0.761) ( 0.068) Fiscal adjustmentt+1 0.000 0.305 0.030 0.301 0.000 0.000 –6.509*** 0.021 x GDP growth ratet+1 … ( 0.328) ( 0.050) (0.129) … … ( 0.208) ( 0.014) Observations 15 81 81 38 8 8 16 81 p-value ... 0.217 0.010 0.014 ... ... ... 0.528 R-squared (within) 0.757 0.049 0.139 0.280 1.000 0.695 0.327 0.002 … = no value; GDP = gross domestic product. Note: Fixed-effects estimation with country-fixed effects included. Robust standard errors in parentheses. *** = p<0.01, ** = p<0.05, * = p<0.1. Source: Authors’ calculations. Appendix 27 Table A7: Past Fiscal Stimuli, Growth, and Development Outcomes (1) (2 ) (3 ) (4 ) (5 ) (6 ) (7 ) (8 ) (9 ) Povertyt UndernourishmenttMortalitytEnrollmenttWomen Powert SanitationtFuel-TechtNon-EducationtBroadbandt GDP growtht–1 1.469** 0.155 0.202* –0.123 –0.535 –0.674 –0.523 0.822 –0.650* ( 0.437) ( 0.178) ( 0.064) ( 0.912) ( 0.479) ( 0.325) ( 0.401) ( 0.627) ( 0.245) Fiscal stimulit–1 4.031* 0.276 0.436 5.767 –0.017 –1.754 –0.523 1.676 –1.546 ( 1.610) ( 0.606) ( 0.361) ( 7.248) ( 2.286) ( 0.851) ( 0.706) ( 1.469) ( 1.365) Fiscal stimulit–1 –1.294 –0.134 –0.237* –2.011 –0.417 0.609 0.221 –0.623 0.825 x GDP growth ratet–1 ( 1.038) ( 0.171) ( 0.075) ( 1.784) ( 0.517) ( 0.285) ( 0.228) ( 0.404) ( 0.381) Observations 58 68 64 74 81 68 64 65 69 p-value 0.101 0.537 0.022 0.171 0.607 0.345 0.463 0.002 0.011 R-squared (within) 0.143 0.036 0.081 0.051 0.068 0.108 0.043 0.112 0.056 (10) (11) (12) (13) (14) (15) (16) (17) Bottom Consumptiont Pollutiont Carbon Dioxide DamagetDisplacementtMarinetTerrestrialtBriberytRemittances GDP growtht–1 -0.134 0.449** –0.056 0.028 –2.699 0.328 0.541 –0.006 (0.143) (0.101) (0.043) (0.063) (3.242) (0.418) (0.364) (0.024) Fiscal stimulit–1 –0.361 2.821** –0.640 0.819 0.000 0.000 1.178 0.051 (0.181) (0.605) (0.291) (0.450) … … (1.554) (0.179) Fiscal stimulit–1 0.000 –0.413 0.137 –0.133 0.000 0.000 0.970 –0.058 x GDP growth ratet–1 … (0.217) (0.084) (0.100) … … (0.434) (0.031) Observations 16 73 77 46 12 12 18 81 p-value ... 0.049 0.154 0.431 0.466 0.490 ... 0.039 R-squared (within) 0.056 0.247 0.114 0.309 0.142 0.015 0.786 0.035 … = no value; GDP = gross domestic product. Note: Fixed-effects estimation with country-fixed effects included. Robust standard errors in parentheses. ** = p<0.05, * = p<0.1. Source: Authors’ calculations. 28 Appendix Table A8: Past Fiscal Adjustment, Growth, and Development Outcomes (1) (2 ) (3 ) (4 ) (5 ) (6 ) (7 ) (8 ) (9 ) Povertyt UndernourishmenttMortalitytEnrollmenttWomen Powert SanitationtFuel-TechtNon-EducationtBroadbandt GDP growtht–1 1.083** 0.144 0.172** –1.070 –0.605 –0.560 –0.476 0.740 –0.540** (0.222) (0.145) (0.040) (0.709) (0.537) (0.251) (0.340) (0.590) (0.094) Fiscal adjustmentt–1 –2.983 0.187 –1.372* –4.723* –2.418 6.380*** 12.408*** –0.965 5.753* (1.888) (0.359) (0.505) (1.534) (5.867) (0.711) (1.370) (2.046) (2.251) Fiscal adjustmentt–1 0.526 –0.106 0.142* 2.793* 0.354 –0.914*** –2.033** –0.107 –0.574 x GDP growth ratet–1 (0.714) (0.126) (0.051) (1.151) (1.049) (0.149) (0.350) (0.617) (0.265) Observations 58 68 64 74 81 68 64 65 69 p-value 0.000 0.236 0.000 0.061 0.328 0.004 0.001 0.062 0.001 R-squared (within) 0.116 0.036 0.081 0.102 0.065 0.117 0.088 0.110 0.077 (10) (11 ) (12 ) (13 ) (14 ) (15 ) (16 ) (17) Bottom Consumptiont Pollutiont Carbon Dioxide DamagetDisplacementtMarinetTerrestrialtBriberytRemittancest GDP growtht–1 0.035 0.304** –0.004 –0.042 –1.615 0.386 0.610** –0.018 (0.369) (0.083) (0.014) (0.076) (2.354) (0.510) (0.135) (0.018) Fiscal adjustmentt–1 –2.321 0.785** –0.179 –0.294 6.787** 0.363 –5.896*** –0.073 (0.875) (0.241) (0.091) (0.367) (1.927) (0.418) (0.598) (0.050) Fiscal adjustmentt–1 0.000 –0.228 –0.036 0.037 0.000 0.000 0.949*** 0.022 × GDP growth ratet–1 … (0.124) (0.022) (0.079) … … (0.162) (0.017) Observations 16 73 77 46 12 12 18 81 p-value … 0.004 0.000 0.671 … … … 0.026 R-squared (within) 0.361 0.118 0.061 0.054 0.294 0.018 0.296 0.009 … = no value; GDP = gross domestic product. Note: Fixed-effects estimation with country-fixed effects included. Robust standard errors in parentheses. *** = p<0.01, ** = p<0.05, * = p<0.1. Source: Authors’ calculations. Appendix 29 Table A9: Large Fiscal Impulses and Tax Components Fiscal StimulitFiscal Adjustmentt (1 ) (2 ) (3 ) (4 ) (5 ) (6 ) (7 ) (8 ) Personal income taxt+1 0.058 0.039 ( 0.049) (0.073) Corporate income taxt+1 0.003 0.048 ( 0.056) (0.058) Other direct taxest+1 –0.242 –0.383 ( 0.220) (0.314) Direct taxest+1 0.025 0.046 ( 0.050) (0.041) Indirect taxest+1 –0.077 –0.072 0.011 0.008 ( 0.055) ( 0.049) (0.036) (0.035) Social contributionst+1 –0.218** –0.232** –0.018 –0.033 ( 0.066) ( 0.060) (0.014) (0.017) Personal income taxt–1 0.019 –0.054 (0.037) (0.030) Corporate income tax t–1 0.095*** –0.039 (0.007) (0.036) Other direct taxes t–1 –0.225*** –0.632*** (0.027) (0.086) Direct taxes t–1 0.067** –0.043 (0.018) (0.045) Indirect taxes t–1 0.031 0.015 –0.007 –0.014 (0.092) (0.076) (0.060) (0.064) Social contributions t–1 –0.089** –0.087** –0.103* –0.135** (0.020) (0.017) (0.037) (0.035) Observations 77 77 81 81 77 77 81 81 p-value ... 0.010 ... 0.000 ... 0.184 ... 0.099 R-squared (within) 0.142 0.124 0.137 0.097 0.054 0.025 0.124 0.059 … = no value. Note: Explanatory variables are cyclically adjusted. Fixed effects estimation with country-fixed effects included. Robust standard errors in parentheses. *** = p<0.01, ** = p<0.05, * = p<0.1. Source: Authors’ calculations. 30 Appendix Table A10: Large Fiscal Impulses and Expense Components Fiscal StimulitFiscal Adjustmentt (1) (2) (3) (4) (5) (6) (7) (8) Waget+1 0.066 0.020 (0.043) ( 0.031) Nonwaget+1 0.065 –0.059 (0.031) ( 0.056) Government consumption expendituret+1 0.066 –0.016* (0.036) ( 0.006) Subsidiest+1 –0.018 –0.018 –0.070 –0.083 (0.037) (0.034) ( 0.060) ( 0.048) Transfert+1 –0.071 –0.071 –0.060* –0.073** (0.049) (0.048) ( 0.022) ( 0.015) Waget–1 –0.020 0.042* (0.046) (0.016) Nonwage t–1 0.002 0.003 (0.042) (0.017) Government consumption expenditure t–1 –0.010 0.025 (0.024) (0.016) Subsidies t–1 0.029 0.031 –0.009 –0.013 (0.050) (0.054) (0.025) (0.027) Transfer t–1 –0.018 –0.012 –0.071 –0.080 (0.039) (0.045) (0.042) (0.047) Observations 77 77 81 81 77 77 81 81 p-value ... 0.292 … 0.285 … 0.061 ... 0.253 R-squared (within) 0.059 0.059 0.009 0.008 0.084 0.074 0.029 0.026 … = no value. Note: Explanatory variables are cyclically adjusted. Fixed effects estimation with country-fixed effects included. Robust standard errors in parentheses. ** = p<0.05, * = p<0.1. Source: Authors’ calculations. Appendix 31 Table A11: Tax/Expense Components and Fiscal Stimuli/Adjustment (1) (2) (3) (4) (5) (6) (7) (8) (9) (10) PITt PITtCITtCITtOthertOthert INDTtINDTtSSCtSSCt Fiscal stimulit–1 0.267 0.025 –0.051* –0.331 –0.408 (0.344) (0.249) (0.016) (0.300) (0.412) Fiscal adjustmentt–1 0.133 0.423 –0.069 0.052 –0.056 (0.167) (0.451) (0.039) (0.176) (0.055) Observations 77 77 77 77 77 77 77 77 77 77 p-value 0.494 0.484 0.926 0.418 0.050 0.177 0.350 0.788 0.395 0.383 R-squared (within) 0.013 0.003 0.000 0.023 0.016 0.031 0.024 0.001 0.096 0.002 (11) (12) (13) (14) (15) (16) (17) (18) Waget WagetNon_wagetNon_wagetSubsidiestSubsidiestTransfer_hattTransfer_hatt Fiscal stimulit–1 0.512* 0.396 –0.131 0.036 (0.183) (0.169) (0.205) (0.163) Fiscal adjustmentt–1 –0.158 –0.593 –0.282 –0.535* (0.117) (0.357) (0.356) (0.195) Observations 77 77 77 77 77 77 77 77 p-value 0.068 0.272 0.102 0.195 0.568 0.486 0.839 0.071 R-squared (within) 0.037 0.004 0.022 0.051 0.005 0.022 0.000 0.049 Note: Dependent variables are cyclically adjusted. Fixed effects estimation with country-fixed effects included. Robust standard errors in parentheses. * = p<0.1. Source: Authors’ calculations. REFERENCES Alesina, Alberto, and Silvia Ardagna. 2010. “Large Changes in Fiscal Policy: Taxes versus Spending.” Tax Policy and the Economy 24 (1): 35–68. Alesina, Alberto, Roberto Perotti, Francesco Giavazzi, and Tryphon Kollintzas. 1995.“Fiscal Expansions and Adjustments in OECD Countries.” Economic Policy 10 (21): 205–48. Blanchard, Olivier Jean. 1990. “Suggestion for a New Set of Fiscal Indicators.” Economics Department Working Paper No. 79. Organisation for Economic Co-operation and Development, Paris. Chongvilaivan, Aekapol, and Annette Chooi. 2021. A Comprehensive Assessment of Tax Capacity in Southeast Asia. Manila: Asian Development Bank. The Growth Lab at Harvard University. 2019. Growth Projections and Complexity Rankings. Harvard Dataverse. doi:10.7910/DVN/XTAQMC. ASIAN DEVELOPMENT BANK ASIAN DEVELOPMENT BANK 6 ADB Avenue, Mandaluyong City 1550 Metro Manila, Philippines www.adb.org Large Fiscal Episodes and Sustainable Development Some International Evidence The link between fiscal policy and sustainable development is country specific. Tax revenues remain a major source of green and social finance. Analysis of episodes of fiscal expansion and tightening in four emerging markets finds that the association between fiscal aggregates and domestic resource mobilization and sustainable development differs across countries. Further, the time paths of tax and expenditure components are important for domestic resource mobilization to target sustainable development outcomes. More granular fiscal data may shed more light on the association between fiscal conditions and sustainable development in the coming years. About the Asian Development Bank ADB is committed to achieving a prosperous, inclusive, resilient, and sustainable Asia and the Pacific, while sustaining its efforts to eradicate extreme poverty. Established in 1966, it is owned by 68 members —49 from the region. Its main instruments for helping its developing member countries are policy dialogue, loans, equity investments, guarantees, grants, and technical assistance. LARGE FISCAL EPISODES AND SUSTAINABLE DEVELOPMENT SOME INTERNATIONAL EVIDENCE Joshua Aizenman, Yothin Jinjarak, Hien Nguyen, and Donghyun Park ADB ECONOMICS WORKING PAPER SERIES NO. 644 December 2021