资源描述：
Policy Research Working Paper 8720 Policy Implications of Non-linear Effects of Tax Changes on Output Samara Gunter Daniel Riera-Crichton Carlos Vegh Guillermo Vuletin Latin America and the Caribbean Region Office of the Chief Economist January 2019 WPS8720 Public Disclosure Authorized Public Disclosure Authorized Public Disclosure Authorized Public Disclosure AuthorizedProduced by the Research Support Team Abstract 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 cutive Directors of the World Bank or the governments they represent. Policy Research Working Paper 8720 An earlier paper titled “Non-linear effects of tax changes on output The role of the initial level of taxation,” estimated tax multipliers using i a novel dataset on value-added taxes for 51 countries 21 industrial and 30 developing for the period 1970–2014, and ii the so-called narra- tive approach developed by Romer and Romer 2010 to properly identify exogenous tax changes. The main finding is that, in line with existing theoretical distortionary and disincentive-based arguments, the effect of tax changes on output is highly non-linear. The tax multiplier is essentially zero under relatively low/moderate initial tax rate levels and more negative as the initial tax rate and the size of the change in the tax rate increase. This companion paper first shows that these findings have important policy implica- tions, given that the initial level of taxes varies greatly across countries and thus so will the potential output effect of changing tax rates. The paper then turns to some specific policy applications. It focuses on the relevance of the argu- ments for revenue mobilization in countries with low levels of provision of public goods and social and infrastructure gaps, as well as in commodity-dependent countries. The paper then considers some practical implications for the standard debt sustainability analysis. Lastly, it uates the implications of the findings for the Laffer curve. 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/ research. The authors may be contacted at gvuletinworldbank.org. Policy Implications of Non-linear E⁄ects of Tax Changes on Output Samara Gunter Colby College Daniel Riera-Crichton World Bank Carlos Vegh World Bank Guillermo Vuletin World Bank JEL Classi cation E32, E62, H20. Keywords tax multiplier, tax policy, tax rate, value-added tax, non-linear, narrative. We would like to thank seminar participants at the International Monetary Fund, Inter-American Development Bank, World Bank, Federal Reserve Board, European Stability Mechanism, Central Bank of Argentina, Central Bank of Chile, Central Bank of Spain, UN Economic Commission for Latin America and the Caribbean, George Washington University, Johns Hopkins University, Williams Col- lege, Davidson College, Universidad Nacional de La Plata, Sao Paulo School of Economics-Getulio Vargas Foundation, Escola Superior d Administraci i Direcci d Empreses ESADE, Institute of Education and Research INSPER, International Macro Workshop-RIDGE, Latin American and Caribbean Economic Association LACEA, Annual Symposium of the Spanish Economic Association, and XLIV Meeting of the Network of Central Banks and Finance Ministries-IDB for many helpful comments and suggestions. We would also like to thank Alberto Alesina, Silvia Albrizio, Leopoldo Avellan, Frank Bohn, Fernando Broner, Eduardo Cavallo, Javier Garcia-Cicco, Aitor Erce, Davide Furceri, Vitor Gaspar, Alejandro Izquierdo, Herman Kamil, Graciela Kaminsky, Aart Kraay, Gerardo Licandro, Alessandro Notarpietro, Peter Montiel, Eduardo Moron, Ilan Noy, Pablo Ottonello, Peter Pedroni, Javier Perez, Roberto Ramos, David Robinson, Diego Saravia, Olena Staveley-O Carroll, Jay Shambaugh, Hamilton Taveras, Teresa Ter-Minassian, and Martin Uribe for helpful discus- sions, and JosØ AndrØe Camarena Fonseca, Diego Friedheim, Pablo Hernando-Kaminsky, and Luis Morano for excellent research assistance.1 Introduction In Gunter et al. 2018, we estimate tax multipliers using i a novel dataset on value-added taxes for 51 countries 21 industrial and 30 developing for the period 1970-2014 and ii the narrative approach developed by Romer and Romer 2010 to properly identify exogenous tax changes. The main empirical nding is that there are strong non-linear e⁄ects of taxation on output. In particular, and in line with existing theoretical distortionary and disincentive-based arguments, we show that the e⁄ect of tax changes on output is essentially zero for relatively low initial tax rate levels or small tax changes and becomes increasingly negative under higher initial tax rates or larger tax changes. The reason is that the distortion imposed by taxation on economic activity is directly, and non-linearly, related to both the initial level of the tax rate and the size of the tax change. Figure1reproducesakeyresultfromourpaper. This gurereportstheestimatedtaxmultiplierafter two years, uated at alternative initial tax rate levels and di⁄erent sizes of the tax change. We can see that the most negative multipliers occur for the highest values in both axes. In other words, while the fall in output associated with increasing revenues by 1 tends to be zero for low levels of the initial tax rate and small tax hikes, this cost becomes much larger as the initial tax rate and the size of the tax hike increase. Hence, the evidence shows that the output e⁄ect of tax changes is, indeed, highly non-linear. Figure 1. Role of initial tax rate level and size of tax rate change Notes Darkbluerepresentsastatisticallyzerotaxmultiplier. Source Gunter, Riera-Crichton, Vegh, andVuletin2018. 2These ndings have important policy implications given that the initial level of taxes varies greatly across countries and thus so will the potential output e⁄ect of changing tax rates. Figure 2 shows that, givencountries currentVATrate, thetaxmultipliercouldbestatisticallyzerodarkbluecolor, or moderate to high yellow, orange, and red colors. For example, a 1.5-percentage-point increase in the VAT rate would essentially not a⁄ect GDP in countries with low tax rates such as Angola, Costa Rica, Guatemala, Ecuador, Nigeria, and Paraguay. In contrast, the same tax increase decrease would cause output to fall increase in economies with relatively high VAT rates, including some emerging markets like Argentina and Uruguay and, especially, many industrial European countries. For example, according to our estimates, the 1-percentage-point increase that took place in Greece in June 2016 when the VAT rate changed from 23 to 24 percent should, in and on itself, have reduced GDP by about 1.75 percent by June 2018. In contrast, the 2-percentage-point increase that took place in Ecuador in January 2000 when the VAT changed from 10 to 12 percent should not, in and on itself, have a⁄ected GDP. Figure 2. Tax multipliers for countries around the world Notes Tax multipliers are calculated based on a 1.5-percentage-point change in the VAT rate. Light blue indicates statistically zero tax multipliers. Source Gunter, Riera-Crichton, Vegh, and Vuletin 2018. Thiscompanionpaperdiscussessomerelevantpolicyapplicationsthatresultfromthenon-lineare⁄ects of tax changes on output. We rst focus on the relevance of our arguments for revenue mobilization in countries with low levels of provision of public goods and social and infrastructure gaps Section 2 as well as in commodity-dependent countries Section 3. Section 4 then explores some practical implications for debt sustainability analysis DSA. Section 5 analyzes the non-linear implications in terms of the La⁄er curve. Finally, Section 6 o⁄ers some concluding remarks. 32 PolicyimplicationsI Ontherelationshipbetweenthesizeofthe government and economic development The role and the size of the government in an economy have been studied from di⁄erent perspectives in the literature. Some of the main determinants of government spending as a proportion of GDP include trade openness Rodrik, 1998, country size Alesina and Wacziarg, 1998, degree of economic development Wagner, 1883, 1893; Easterly and Rebelo, 1993, political organization Persson and Tabellini, 1999; Milesi-Ferretti, Perotti, and Rostagno, 2002, and business cycle volatility Fatas and Mihov, 2001. Within this broad set of theories, one that has received considerable attention focuses on the relation- ship between the size of the government and the degree of economic development. In fact, one of the best-established empirical regularities in public nance is the existence of a positive relationship between the size of government spending relative to GDP and real GDP per capita. This so-called Wagner s Law, also known as the law of increasing state activities, is named after the German econo- mist Adolph Wagner s empirical analysis of Western Europe at the end of the 19th century. Wagner 1883, 1893 pointed out that, as national income rises, public spending tends to increase, both at the extensive margin i.e., new activities are undertaken and at the intensive margin i.e., existing activities are pered on a larger scale. Speci cally, Wagner argued that the public sector would take over and expand administrative, regulatory, and protective activities previously pered by the private sector because, as nations develop, they face increased complexity in terms of legal and administrative frameworks. Wagner also predicted the expansion of public expenditures related to culture and welfare based on the presumption that, as income rises, citizens will increase the demand for services, such as education, public health, old age pension, and other social protection programs. In fact, these types of goods and services are viewed as luxury goods i.e., the income elasticity of demandexceedsunityandhave, inprinciple, morecharacteristicsofpublicgoodsthanprivategoods, in which case public provision seems to be the norm. Finally, Wagner was of the view that govern- mentinterventionwouldberequiredtomanageand nancenaturalmonopoliesandensurethesmooth operation of market forces. While not uncontested, these arguments were later re ned by, among others, Peacock and Wiseman 1967, Musgrave 1969, and Bird 1971. Figure 3 shows that, as predicted by Wagner s law and a large, and more recent, body of empirical evidence,thereisindeedaverystrongassociationbetweenGDPpercapitaandthesizeofgovernment spending relative to GDP. 1 Based on this framework, countries above below the tted line are 1 Most of the recent empiricalevidence regarding the existence of Wagner s law is based on country-speci c analyses as opposedtocross-sectionalanalysesliketheoneshowninFigure3. While, strictlyspeaking, across-sectionalanalysisis technicallylessaccurateasitdoesnotcontrolforothercountrydeterminantsthatmaybeconstantovertime, thismore 4countries with a size of government spending larger smaller than that of a typical country with the same level of income per capita. 2 For example, Honduras a lower-middle-income country, based on the World Bank income classi cation, with a GDP per capita of 4,785 has a ratio of public spending to GDP of 23 percent, while Jamaica s an upper-middle-income country with a GDP per capita of 8,528 is 27 percent. However, Costa Rica an upper-middle-income country with a GDP per capita of 14,471 i.e., a per capita income 3 times as high and 70 percent larger than that of Honduras and Jamaica, respectively has a ratio of public spending to GDP of just 18 percent. Figure 3. Relationship between GDP per capita and size of government spending relative to GDP Notes Data correspond to the year 2015. Total number of countries in the sample is 107. Sources Data for government spending over GDP ratio from WEO-IMF. Data for GDPpc from Penn World Tables. Interestingly, de ning excess spending as the ratio G/GDP minus the predicted ratio G/GDP from the tted line in Figure 3, it follows, from individual country examples in Figure 3 and more system- atic evidence from Figure 4, that countries with positive excess spending like Honduras, Jamaica, Argentina, and Greece tend to have higher VAT rates than those with negative excess spending like Guatemala, Costa Rica, New Zealand, and Australia. global cross-section analytical framework should allow us to draw some relevant policy insights in a more transparent way. 2 For comparison purposes, GDP per capita is PPP-based. Figures are in constant 2011 PPP dollars. 5Figure 4. Relationship between excess spending and VAT rate Notes Data correspond to the year 2015. Total number of countries in the sample is 107. Sources Data for government spending over GDP ratio from WEO-IMF. Data for GDPpc from Penn World Tables. This evidence suggests that countries such as Guatemala, with a low level of provision of public goods for its degree of development, may reach a more typical level of provision of public goods by collecting more revenues from increases in the VAT rate with little e⁄ect on economic activity. Note that this analysis is quite conservative regarding the positive e⁄ect of these scal changes since it does not include the potential output e⁄ect of higher government spending associated with larger scal revenues. In fact, in the 2016 Article IV Consultation for Guatemala, the IMF argued that [w]hilethe rst-bestoptionistoquicklymobilizerevenuetocovertheoriginallyplanned budget spending in 2016, a temporary relaxation of the overall de cit to its historical averageof2percentofGDPcouldbejusti edtopreventfurthercutsinsocialandcapital spending in the event of revenue shortfalls. The acute problem of extreme poverty and malnutritionaswellasthedireinfrastructureandsecurityneedscouldwarrantmaintaining a de cit at 2 percent of GDP, or even modestly higher, over the medium term to allow the time needed for revenue mobilization. This would not jeopardize scal sustainability AnnexIIIandbeconsistentwithcurrentimplementationcapacity.[Moreover,]taxpolicy measures are also needed. Re of tax and customs administration is a key prior
展开阅读全文