How do the tax burden and the fiscal space in Latin America look like? Evidence through Laffer Curves

Lozano-Espitia, Ignacio; Arias-Rodríguez, Fernando Article How do the tax burden and the fiscal space in Latin America look like? Evidence through Laffer Curves Latin American Economic Review Provided in Cooperation with: Centro de Investigación y Docencia Económica (CIDE), Mexico City Suggested Citation: Lozano-Espitia, Ignacio; Arias-Rodríguez, Fernando (2021) : How do the tax burden and the fiscal space in Latin America look like? Evidence through Laffer Curves, Latin American Economic Review, ISSN 2196-436X, Centro de Investigación y Docencia Económica (CIDE), Ciudad de México, Vol. 30, Iss. 1, pp. 1-26, https://doi.org/10.47872/laer.v30.77 This Version is available at: https://hdl.handle.net/10419/259467 Standard-Nutzungsbedingungen: Die Dokumente auf EconStor dürfen zu eigenen wissenschaftlichen Zwecken und zum Privatgebrauch gespeichert und kopiert werden. Sie dürfen die Dokumente nicht für öffentliche oder kommerzielle Zwecke vervielfältigen, öffentlich ausstellen, öffentlich zugänglich machen, vertreiben oder anderweitig nutzen. Sofern die Verfasser die Dokumente unter Open-Content-Lizenzen (insbesondere CC-Lizenzen) zur Verfügung gestellt haben sollten, gelten abweichend von diesen Nutzungsbedingungen die in der dort genannten Lizenz gewährten Nutzungsrechte. Terms of use: Documents in EconStor may be saved and copied for your personal and scholarly purposes. You are not to copy documents for public or commercial purposes, to exhibit the documents publicly, to make them publicly available on the internet, or to distribute or otherwise use the documents in public. If the documents have been made available under an Open Content Licence (especially Creative Commons Licences), you may exercise further usage rights as specified in the indicated licence. https://creativecommons.org/licenses/by-nc-nd/4.0/ Latin American Economic Review Open Access RESEARCH ARTICLE Lozano and Arias, Lat Am Econ Rev (2021) 30:9 https://doi.org/10.47872/laer.v30.77 How do the Tax Burden and the Fiscal Space in Latin America look like? Evidence through Laffer Curves Ignacio Lozano-Espitia1; Fernando Arias-Rodríguez2 1 Banco de la República de Colombia. E-mail: [email protected] 2 The authors are Director and Expert Professional of the Economic Policy Studies Department, DEPE, of the SGEE at Banco de la República, (the Central Bank of Colombia), respectively. Postal address: Carrera 7 # 14-78, Bogotá – Colombia. The points of view expressed in the paper belong to the authors only, and do not represent those of Banco de la República or its Board of Directors. Correspondence: [email protected] Abstract How much fiscal space do Latin American countries have to increase their tax burdens in the long term? This paper provides an answer through Laffer curves estimates for taxes on labor, capital, and consumption for the six largest emerging economies of the region: Argentina, Brazil, Chile, Colombia, Mexico, and Peru. Estimates are made using a neoclassical growth model with second-generation human capital and employing data from the national accounts system for the period from 1994 to 2017. Our findings allow us to compare the recent effective tax rates on factor returns against those which would maximize the government’s revenues, and therefore to derive the potential tax-related fiscal space. Results suggest that joint fiscal space on labor and capital taxes would reach 6.5% of GDP for the region, on average, although there are important differences among the countries. Classification JEL: E13, E62, H20, H30, H60 Key Words: Laffer curves; fiscal policy; taxes on consumption, taxes on labor and capital income. How do the Tax Burden and the Fiscal Space in Latin America look like? Evidence through Laffer Curves Lozano and Arias 2/26 Introduction Professor Arthur Laffer from the University of Southern California made the following assertion in 1974: “...there are always two tax rates that yield the same revenues…” (Wanniski, 1978). To explain his statement, he drew an inverted u-shaped curve with the government revenue (on the vertical axis) increasing with the tax rate (on the horizontal axis) until a maximum point, where it begins to decrease. Since then, the simple design of the Laffer Curves has crucial implications, given that an increase in taxes could have two opposite effects regarding fiscal revenues: On the one hand, an increase; on the other, a reduction, because a higher tax could disincentivize the labor supply and the desire to invest. One of the two effects will predominate in an economy depending on the tax burden level with respect to its maximum point. If it stands below, increasing taxes could generate more revenues. However, if it stands above, increasing taxes could end up reducing government revenues. Therefore, each curve shows how government revenues at steady state vary when there are changes on the effective labor tax rate, while keeping the other tax rates and the remaining parameters constant. Hence, policymakers are usually concerned with knowing the rates that maximize their country’s fiscal revenues and therefore how close/far the current tax burden is with respect to such maximizing rates. In this paper, we estimate the Laffer Curves for taxes on labor, capital, and consumption expenditures for the six largest Latin American emerging economies: Argentina, Brazil, Chile, Colombia, Mexico, and Peru. Estimates are made employing a neoclassical growth model with human capital accumulation, where growth rates are defined both exogenously and endogenously, as suggested by the literature (Trabandt and Uhlig, 2011; Lucas 1988; Uzawa, 1965). The model is calibrated with annual data coming from the national accounts of each country (United Nations system) for the period from 1994 to 2017. The effective tax rates—one of the most important parameters—are calculated following the procedure applied by Mendoza et al. (1994) to G7 countries, although we include some extensions considered by Prescott (2004) and Trabandt and Uhlig (2012). The long-term government’s potential tax-related fiscal space is the result of contrasting effective tax rates on labor and capital income (tax burden) against those that would maximize the government’s revenues (obtained from the model). By comparing the effective tax rate that maximizes labor income versus the most recent rate, results suggest that governments in Latin American Countries (LAC) would have an important labor tax-related fiscal space. It reaches 13.7%, on average, for the region, meaning that for every 100 bp (basic points) of increase in the effective tax rate on labor, the tax collection on labor income would increase 44 bp. Regarding the Laffer curves associated to capital taxes, figures suggest that fiscal space is smaller. Thus, on average, the regional space would reach 10% of capital income, meaning that for every 100 bp of increase in the effective tax rate, the capital tax collection would increase only by 25 bp. As a share of potential GDP, fiscal space on labor and capital incomes achieves 3.7% and 2.7%, respectively. Adding up the figures on labor and capital Laffer Curves, results suggest that fiscal space could achieve nearly 6.5% of potential GDP, on average for the region, which gives an important margin to adjust the tax burden in the long term. This percentage would be larger if we add the potential space from the consumption taxes. Of course, there are important differences in fiscal space for both factors when they are assessed and compared between countries. Mexico and Chile seem to have the highest fiscal space, while Brazil and Argentina the lowest. To the best of our knowledge, this is the first time that a human capital growth model has been considered to estimate Laffer curves and the fiscal space for the main Latin American emerging economies Latin american economic review (2021) 30:9 3/26 The paper contains six sections in addition to this brief introduction. Section 2 presents the theoretical model of human capital with exogenous accumulation, which we use as baseline and is labeled as 2nd generation. In section 3 we describe the calibration and parameterization of the model, with special mention to the computation of effective tax rates. Section 4 presents the main results, and in section 5 we compare our results with respect to those coming from an endogenous human capital model. Section 6 contrasts the findings for Latin America on tax burden and fiscal space with respect to developed countries (USA and the 14 main European Union economies, or 14-EU). Finally, Section 7 presents some final remarks. We include five appendices with methodological details on the effective tax rates calculation and with supplementary country specific as well as aggregated results that help us in explaining what we find on section 4. The Theoretical Approach Agents There are three agents: a representative household, a representative firm, and a government financed with distortionary taxes and debt to provide public goods and transferences to households. At the beginning of each period, the household has assets represented by the stock of physical capital, , human capital, , and government bonds, . The household rents capital to the firm which is used to produce final goods and receive a rental rate given by where is the depreciation rate of physical capital. The household also obtains interest payments, , for holding government bonds. Additionally, households obtain a wage, , for the hours hired in the labor market, profits from firms, , since they are their owners, transferences from the government, , and other exogenous resources denoted by . The latter could be associated to transferences from abroad. Consequently, this economy could be modeled as an open economy without further complications.1 The model used in this paper includes human capital both exogenously and endogenously accumulated. Thus, the hours offered by the households to the market depend on their desire to increase their human capital (in the form of studies and/or learning-by-doing). This is a key feature that makes our theoretical approach different from the standard neoclassical model of exogenous growth used as benchmark by Trabandt and Uhlig (2011) and Nutahara (2015) in previous papers on this matter. The labor tax rate charged by the government, , implies, therefore, that it could affect not only labor-leisure household decisions, but also the decisions between learning hours and those effectively worked for the firm. Households also pay a tax, , on capital returns. From the perspective of disposable income, households choose how much to spend in consumption and how much to accumulate in assets. Furthermore, the government imposes a tax rate to consumption given by . The representative household maximizes the discounted value of its utility (equation 1) subject to a budget constraint (equation 2), and to the equations representing the law of motion for physical and human capital (equations 3 and 4): (1) Subject to: 1 Although it is a very simple form of modelling an open economy. How do the Tax Burden and the Fiscal Space in Latin America look like? Evidence through Laffer Curves Lozano and Arias 4/26 (2) (3) (4) where represent consumption, hours worked, investment in physical capital, government bonds, government spending, transfers, and stock of physical and human capital, respectively. As mentioned, households increase their human capital through studying and/or learning by doing. This implies that they devote a fraction of their time, , to increase their knowledge and the remaining portion, , to work at the firm. Human capital depreciates at a rate of each period. The motion of human capital accumulation follows the ideas developed by Levine et al. (1992). Household preferences are represented by equation (5), where is the inverse of intertemporal rate of substitution, measures the labor disutility, and represents Frisch elasticity (sensitivity of labor supply to changes in the disposable labor income) that could be affected by tax changes. The way in which the model includes preferences is standard in all the previously cited papers. (5) In turn, the representative firm operates under perfect competition and uses capital and labor as inputs to produce a homogeneous final good . Firms maximize their profits, which are given by: (6) subject to a Cobb-Douglas production function: (7) where represents the total factor productivity. The work effectively used by firms, , is affected by the workers’ training, . The government faces the following budget constraint: (8) Equation (8) shows that in every period, public spending plus transferences to households plus the debt service must be equal to revenues from new debt issued by the government, whose maturity is equal to one period, plus taxes charged to factors and consumption. The total tax revenue, , is given by . The first order conditions for both the representative household and the firm and details on the steady-state growth solution can be consulted in Lozano and Arias (2018). Latin american economic review (2021) 30:9 5/26 Equilibrium and Characterization of Laffer Curves The model´s equilibrium follows the insights provided in Trabandt and Uhlig (2011). All variables - except hours worked, interest rate, and tax rates - grow at a constant rate equal to: (9) The optimal capital-output ratio consistent with the balanced growth path is given by: (10) whereas the optimal work level under such specification is: (11) where (12) Notice that fraction of hours offered by a household depends on the consumption-output ratio multiplied by a term which, in turn, depends on taxes on consumption and labor, the Frisch elasticity and the variables related to human capital accumulation as well. This result comes essentially from if the same forces that govern growth, via physical capital accumulation, dominate long-term growth. In addition, the model assumes that human capital greatly affects the labor supply through its accumulation. The human capital stock at the balanced growth path is defined as a working time fraction offered by a household, so that: (13) In order to characterize the Laffer Curves,2 it is assumed that both debt and government spending are aligned with the balanced growth path, i.e., and , whereas the transferences to households are equal to the residual needed to meet the government budget constraint: . Hence, the higher tax revenues are for the government, the greater the transferences to households. These remarks are consistent with the ideas by Levine and Renelt (1992), who argue that little evidence is found concerning a tax system affecting the long-run growth rate [also discussed by Trabandt & Uhlig 2011]. Hence, taxes in the model impact economy only through the level of human capital, and since human capital is proportional to non-leisure time along the balanced growth path, there seems to be no intertemporal effect, as is the case with physical capital. Because of this, the approach is commonly known as human-capital-exogenous-growth 2 The Laffer Curves displayed in Section 4 come from a one-dimensional nonlinear equation that depends mainly on the labor supply. It is solved numerically using values given in Table 1 and changing the corresponding tax rate across all possible values it can take. How do the Tax Burden and the Fiscal Space in Latin America look like? Evidence through Laffer Curves Lozano and Arias 6/26 model (our benchmark model). In section 5 we include results from the endogenous version, i.e., a model that allows for long-term growth impacts of taxation. Calibration and Parameterization Effective Tax Rates A key parameter to estimate Laffer Curves is the tax rates. As described in section 2, the government sets three taxes: one on labor income, , another on earnings from renting physical capital, , and the last on consumption of goods and services, . Empirical literature favors the use of effective tax rates instead of statutory rates, since they can better reflect the real tax burden. Effective tax rates are usually lower than statutory rates mainly due to tax evasion and tax benefits contemplated in the tax codes. Speaking broadly, an effective tax rate is defined as the tax collection as a percentage of an amount of income (or expenditure) susceptible of being taxed, known as the tax base. Methodologically, these rates are calculated in this paper following the procedure applied by Mendoza et. al. (1994) to G7 countries with some extensions considered by Prescott (2004) and Trabandt and Uhlig (2012). Effective tax rates are calculated using annual data from the National Accounts (NA) database of each country from 1994 to 2017 (except for Brazil because the information is only available since 2000). We also use information from the OECD and Penn World Tables databases. The NA tax records include revenues from the central government as well as subnational governments. Labor taxes consider all the households’ labor earnings (including the so-called mixed income) as well as social security contributions and payroll taxes. Regarding taxes set on capital gains, they include the taxes paid by corporations as well as those paid by households that hold businesses. In Appendix 1, we present further details on these tax rate estimates. Results are shown in Graph 1. Graph 1. Effective Tax Rates in Latin American Countries A. On Labor Income Latin american economic review (2021) 30:9 7/26 Graph 1 (continued). Effective Tax Rates in Latin American Countries. B. On Capital Gains C. On Consumption Expenditures Source: Calculations by the authors based on data from the National Accounts. Graph 1 includes the effective tax rates for the six LAC and the weighted average of the region (dotted line), using the size of the tax basis as weighting. As we can see, there is an ample difference in tax burden among these six economies. Nonetheless, when dynamics during the last twenty-five years are examined carefully, such differences have been reducing. On labor earnings, for instance (Panel A), the highest effective tax rate is Brazil’s (31% for 2017), while the lowest are Peru’s and Argentina’s (17%). The most recent difference of 14 percentage points (pp) between these countries is lower than those at the beginning of the century (20 pp). The regional weighted average is about 25%, being Argentina, Colombia, and Chile the countries with the largest effort to increase the tax burden on labor in recent times. Trends are not different for taxes on capital gains. The highest effective tax rate is again in Brazil (21% in 2017, close to Argentina) and the lowest is for Peru and Mexico (11.8%). The current difference in capital tax burden between these economies is only of 9 pp, while ten years ago it was close to 17 pp. On average, the region currently has a 17.5% tax burden on capital returns, while in the mid-nineties such tax burden was 7.5%, clearly denoting important advances How do the Tax Burden and the Fiscal Space in Latin America look like? Evidence through Laffer Curves Lozano and Arias 8/26 on this matter. It is remarkable that for both labor and capital incomes, Brazil has the highest effective tax rates, although the latter has been importantly reduced in recent years (falling from 27% to 19% between 2007 and 2014). In turn, Mexico appears to have the least expensive tax system for production factors.3 Regarding consumption, the weighted average tax rate for the region was 19% in 2017, being the highest rate for Argentina (31.5%) and the lowest for Mexico (8.6%). Argentina’s effort to increase consumption tax revenues is evident especially since the beginning of this century. The differences in tax burden on consumption are more ample among countries, and there is less convergence between them. Other Parameters Another set of parameters is required to estimate the Laffer curves. According to their source, we classified them into those calibrated from National Accounts and IMF statistics, other estimates by previous literature, and those calibrated into the model by finding the steady-state solutions. Table 1 summarizes these parameters for each one of the six LAC as well as the average for this group of economies. Following the model notation (Section 2), the share of capital on output (θ), the GDP growth rate (ψ), and the capital-output ratio (K/Y) are averaged values estimated from national accounts, while the current account balance, (m), government expenditures, (g), and public debt, (b), come from the IMF’s World Economic Outlook. The labor supply, (h), is calculated as the average number of hours worked by a worker daily divided by fourteen hours, which are understood as the total number to be distributed between labor and leisure (Ragan, 2006). This data is sourced in the Conference Board. On the second group we include the discount factor, (β), whose value for each country is taken from Lama (2011), the capital depreciation rate, (δ), which is set by using data from Penn World Tables (Feenstra, Inklaar, &Timmer, 2015), and the inverse elasticity of substitution, (η), which is set independently for each country according to their own studies. Finally, Frisch elasticity (φ) is set at 1 (in the benchmark model) as is done by Trabandt & Uhlig (2011), Fernandez-Villaverde et al. (2015), and Chetty et al. (2011).4 In the last group we include government transfers to households, (s), and the steady state gap between the real interest rate and output growth. The former is calibrated by the model as the value that balances the government’s budget constraint, while the latter uses the same real interest rate for all countries - measured as the weighted average of the money market for the six LAC - to avoid extremely positive high gap values, as is the case with Brazil, or negative values as those for Peru. Table 1. Main Macroeconomic Parameters for LAC (Average, 1994-2017) Parameters Argentina Brazil Chile Colombia Mexico Peru LAC Discount Factor 0.954 0.9317 0.921 0.925 0.947 0.936 0.938 Capital Depreciation Rate 0.0346 0.0436 0.0364 0.0384 0.0353 0.0342 0.0390 Inverse of IES 2 2 1.5 2 2 4 2.05 3 See a more detailed estimate of effective tax rates in Brazil and Mexico with Mendoza’s (1994) methodology; annual and quarterly data are provided by Freitas and Marsiglia (2015) and Antón-Sarabia (2005), respectively. 4 Our benchmark model uses this value specially following Keane (2021), who shows that papers that account for human capital or the participation margin usually find Frisch elasticities around 1.0. We acknowledge that there is a lack of consensus about the magnitude of the Frisch elasticity. The literature that employs microeconomic data (for example Park (2020), Kneip et al (2020), Chang et al (2019), Keane & Wasi (2016), and Attanasio et al. (2018)) argues that it is usually less than 1.0. With macroeconomic data, Frisch elasticity could take values from 1.0 to 4.0 (Whalen and Reichling (2016), Fiorito & Zanella (2012), Banerjee (2014), Cacciatone and Traum (2020), Gottlieb et al. (2021)). Latin american economic review (2021) 30:9 15/26 The Laffer Curves displayed in Graph 5 are also estimated using the effective tax rates of 2017 and the parameters from Table 1. In blue, we show curves resulting from the endogenous growth model, while the ones from the exogenous (the baseline model) are shown in red. For labor returns (Panel A), the new effective tax rate that maximizes revenue is to the right on the baseline, implying an increase in tax revenues (fiscal space) of around 18% (resulting from 32% - endogenous - minus 14% - exogenous). In the case of capital returns (Panel B), the opposite happens: the effective tax rate that maximizes revenues is to the left of the baseline. Nevertheless, the fall in tax revenues is marginal, only around 0.9% (9.1% - 10%) (see Tables from graphs 5 and 4). To better understand these results, it is enough to analyze what happens when there is a change in labor tax, first along the Laffer Curves and second at any specific point. Thus, if the labor tax increases, the labor income falls, (1-τ)w, pushing down the non-leisure time, (n), the share of non-leisure time for work, (q), and, therefore, the working time, (q*n). This, in turn, leads schooling time to grow, ((1-q)*n), since the model assumes that the households are always employed. Because the growth is driven by (q*n), there will be less economic growth and less physical capital accumulation, output level, and lower interest rates. The left-side of the capital Laffer Curve (Panel B) is just a consequence of the lower real interest rate, while the insignificant change in capital tax revenue comes from the fact that human capital is the growth source of this economy. Nonetheless, from the second point of view, the greater labor fiscal space from the endogenous model is a consequence of having both larger interest rates and economic growth at a particular labor tax rate. For instance, using the effective tax rates of 2017, we get larger real interest rate and long-run economic growth, than those used for the baseline (6.3% vs. 5.1% and 3.2% vs. 2.7%, respectively). Tax Burden and Fiscal Space: Latin America vs. E.U.-14 and U.S.A. In this section, we contrast Latin America’s Laffer Curves for labor and capital taxes with those corresponding to the main European Economies and the U.S.A. As described in Section 3, we consider the largest six Latin American economies (LAC) and, for the European case, we retake the main fourteen economies (EU-14), as in Trabandt and Uhlig (2011).6 The graphs depict the weighted average for these two groups of economies. The comparison is informative because in all cases we use the same benchmark model (with unitary Frisch elasticity), the data comes from national accounts and covers the same period, and we apply a similar technique to estimate effective tax rates.7 Graph 6 exhibits the results. On the left-hand side, we compare the Laffer curves for labor taxes. The fiscal revenues are standardized, so the value of 100 corresponds to the last available effective tax rate, which is 25% for LAC in 2017 (dotted green vertical line), while for the U.S.A. and the E.U.-14 they are 28% and 46%, respectively (dotted blue and red vertical lines). On the right-hand side, we display the corresponding Laffer curves for capital taxes. 6 The EU-14 group includes Belgium, Denmark, Germany, Ireland, Greece, Spain, France, Italy, The Netherlands, Austria, Portugal, Finland, Sweden, and the United Kingdom. 7 For U.S.A and EU-14 we update the Trabandt and Uhlig´s database for all parameters up to 2017 (using European Commission (2019), and OECD (2019). and re-estimate the models How do the Tax Burden and the Fiscal Space in Latin America look like? Evidence through Laffer Curves Lozano and Arias 16/26 Graph 6. Laffer Curves on Labor and Capital Incomes: LAC vs. EU-14 and & U.S (Exogenous HC and Frisch 1) A. Labor B. Capital Labor Capital Last Tax Rate, 2017 Maxim. Rate Fiscal Space (% labor income) Fiscal Space (% GDP) Last Tax Rate, 2017 Maxim. Rate Fiscal Space (% capital income) Fiscal Space (% GDP) LAC 0.25 0.57 13.71 3.72 0.17 0.57 10.06 2.66 USA 0.28 0.48 8.27 2.09 0.34 0.60 4.10 1.04 EU14 0.46 0.40 - - 0.36 0.29 - - Source: Authors’ calculations. Regarding labor levies, we stress that the current effective tax rate for the E.U.-14 group is over to the one that maximizes the fiscal revenues, implying that additional increments in labor taxes could create disincentives on labor supply, therefore reducing tax revenues. This finding is not surprising, since, as it is well known, European countries such as Sweden (54%), Austria (52%), and France (49%) are among those with the largest labor tax rate in the whole world. In this sense, there seems to be no fiscal space to set more labor taxes in the region. Our results also suggest a little different situation for U.S.A. and LAC. In the former, there is a significant fiscal space (due to the ample distance between the current effective labor tax rate vs. the one that maximizes revenues), which could turn into additional tax revenue close to 8.3% of the labor income. For the latter group such revenues could be around 13.7%. With respect to capital, the right panel of Graph 6 shows that Europe possesses the highest effective tax rate (36%) compared to the U.S.A. (34%) and LAC (17%). The U.S.A.’s effective tax rate corresponds to 2017 and, therefore, does not totally capture the important tax reduction to corporate profits set at the beginning of the Trump administration. Despite this, there would be an important space to increase the U.S.A. tax burden, although with relatively small revenue gains. Thus, hypothetically, should the effective rate be increased to the level that maximizes revenues, gains in tax revenues would be only 4.1% of the capital returns. Europe’s effective tax rate on capital is beyond to the peak of the Laffer Curve, which in practice means that there is no fiscal space. Within the E.U.-14 there are economies such as France (50%), Denmark (47%), Belgium (44%), and United Kingdom (40%) with capital tax burdens that leave narrow or nil space to consider additional rises. In turn, the LAC group has the largest fiscal space due to the smaller current tax burden. Latin american economic review (2021) 30:9 17/26 The degree of concavity for the Laffer Curves, which depends on fundamental parameters of each economy, also determines the size of the fiscal space (the biggest concavity being for the U.S.A. and the lowest for E.U.-14). Conclusions and Final Remarks Laffer Curves estimates are crucial for economic policy actions due to two potential opposite effects that a revenue-seeking tax reform could have: on one side, it could bring an expected increase; on the other, a reduction because higher taxes could disincentivize the labor supply and the desire to invest. The status of the tax burden with respect to the one that generates the maximum revenues determines which of the two effects ends up dominating. The main goal of this paper was to measure how close/far are the current effective tax rates in LAC with respect to those that maximize revenues and, therefore, to explore the potential tax-related fiscal space. The Laffer curves on labor and capital taxes were estimated using a neoclassical growth model with human capital for the six largest Latin American emerging economies: Argentina, Brazil, Chile, Colombia, Mexico, and Peru. The model was calibrated with annual data from the national accounts of each country for the period 1994-2017. To the best of our knowledge, this is the first time that a human capital growth model has been considered to estimate Laffer curves and the fiscal space for the main Latin American emerging economies. There are ample differences across the LAC concerning the recent tax burden, even though such differences have been declining. Regarding labor, Brazil has the highest effective tax rate (31% for 2017), while the lowest is for Peru and Argentina (17%). The regional weighted average is 25%, being Argentina, Colombia, and Chile the countries who have made more efforts to increase these taxes in recent times. Trends are not so different for taxes on capital returns. The highest effective tax rate is again in Brazil (21% in 2017, close to Argentina) and the lowest is for Peru and Mexico (11.8%). On average, the region has now a 17.5% tax burden on capital returns, while in the mid-nineties it was 7.5%. When these effective tax rates are contrasted against those which would maximize the government’s revenues (obtained from the model), a long-term potential tax-related fiscal space is found. Concerning labor taxes, results suggest that governments in the region have an important margin to increase their revenues. In terms of potential GDP, the long-term fiscal space on labor income would reach 3.7%. In detail, the figures suggest that for every percentage point that the effective tax rate increases, the revenues from labor taxes would increase by 0.44 points, on average, for the six countries. The lower the elasticity of the labor supply, the greater the revenue gains. Of course, results differ across countries, being Mexico and Chile those with the largest fiscal space and Brazil with the lowest one. Fiscal space is smaller for taxes levied on capital returns. For this group of LAC, on average, it would achieve 2.7% of the potential GDP. Results also suggest that for every percentage point of increase in the effective tax rate, capital revenues increase 0.25 points. Mexico appears to have the greatest potential space to increase tax revenues on capital, while Peru, Argentina, and Brazil have the lowest. When contrasting the fiscal space for LAC with that of the main European economies (on average), we find that the latter practically do not have margin to increase taxes on neither labor nor capital incomes. By contrast, compared to the U.S.A., results suggest that this country has fiscal space particularly though labor taxes. We conclude with three important remarks which help to understand the actual scope of this paper. First, our analysis is far from being a social welfare analysis since the effective tax rates that maximize revenues are not optimal tax rates. Second, our benchmark model is designed for a closed economy and it does not consider an informal labor market. These issues restrict How do the Tax Burden and the Fiscal Space in Latin America look like? Evidence through Laffer Curves Lozano and Arias 18/26 our analysis because, on the one hand, it does not capture the external competitive context and the recent debates in the region about the need to reduce corporate taxes in order to attract investment and thereby strengthen the taxable bases. On the other hand, the significant degree of informality that Latin American labor markets endure is left aside. Finally, when we use effective tax rates to take advantage of the fiscal space, we have to bear in mind that changes in these effective tax rates can be achieved through an extensive menu. For instance, by including new taxpayers into taxable bases, and/or by reducing or eliminating tax benefits usually granted through tax codes, and/or adjusting statutory tax rates, and/or through better tax management that lead to lower evasion. Acknowledgements The authors wish to thank Gloria Duran, Carolina Arcila, Valerie Carbonell, and Héctor Segura for their valuable assistance. References Airaudo, M., Buffie, E. F. & Zanna, L. F. (2016). “Inflation Targeting and Exchange Rate Management in Less Developed Countries”. IMF Working Paper. Antón-Sarabia, A. (2005). Average Effective Tax Rates in Mexico. Economía Mexicana NUEVA ÉPOCA, XIV, issue 2, p. 185-215. https://econpapers.repec.org/article/emcecomex- /v_3a14_3ay_3a2005_3ai_3a2_3ap_3a185-215.htm Attanasio, O., Levell, P., Low, H., & Sánchez‐Marcos, V. (2018). Aggregating elasticities: intensive and extensive margins of women’s labor supply. Econometrica, 86(6), 2049-2082. Banerjee, R. (2014). An evaluation of the revenue side as a source of fiscal consolidation in high debt economies. Journal of Economic Studies. Brazilian Institute of Geography and Statistics (IBGE) (2019). “Integrated Economic Accounts”. Retrieved from: https://www.ibge.gov.br/en/statistics/economic/national-accounts.html Cacciatone, M., & Traum, N. (2018). Trade flows and fiscal multipliers. The Review of Economics and Statistics, 1-44. Chang, Y., Kim, S. B., Kwon, K., & Rogerson, R. (2019). 2018 Klein lecture: individual and aggregate labor supply in heterogeneous agent economies with intensive and extensive Margins. International Economic Review, 60(1), 3-24. Chetty, R., Guren, A., Manoli, D. & Weber, A. (2011). “Are Micro and Macro Labor Supply Elasticities Consistent? A Review of Evidence on the Intensive and Extensive Margins”. American Economic Review, 101(3), 471-475. Chile’s Central Bank (2019). “Statistics”. Retrieved from: https://www.bcentral.cl/en/web/ banco-central/areas/statistics Conference Board (2019). “Productivity & Innovation-Total Economy Database”. Retrieved from: https://www.conference-board.org/data/economydatabase/index.cfm?id=27762 European Commision. (2019). "eurostat" Retrieved from https://ec.europa.eu/eurostat/web/ main/data/statistics-a-z Feenstra, R. C., Inklaar, R. &Timmer, M. P. (2015). “The Next Generation of the Penn World”. American Economic Review, 105(10), 3150-3182. Fernández, J., Guerrón, P., Kuester, K. & Rubio, J. (2015). “Fiscal Volatility Shocks and Economic Activity”. American Economic Review, 105(11), 3352-3384. Fiorito, R., & Zanella, G. (2012). The anatomy of the aggregate labor supply elasticity. Review of Economic Dynamics, 15(2), 171-187. Latin american economic review (2021) 30:9 19/26 Freitas C., & Marsiglia A. 2015. "Effective Tax Rates on Consumption and Factor Incomes: a quarterly frequency estimation for Brazil". Working Papers Series 398, Central Bank of Brazil, Research Department. https://ideas.repec.org/p/bcb/wpaper/398.html Gómez, M. A. (2016). Are taxes a good predictor of time use patterns? Examining the role of some key elasticities. Economic Modelling, 55, 394-400. Gottlieb, C., Onken, J., & Valladares-Esteban, A. (2021). On the Measurement of the Elasticity of Labour. European Economic Review, 103879. Inter-American Center of Tax Administrations (CIAT) (2019). “Tax Rates in Latin America”. Retrieved from: https://www.ciat.org/tax-rates-in-latin-america/?lang=en International Monetary Fund (IMF) (2019). “World Economic Outlook”, April 2019. Retrieved from: https://www.imf.org/en/Publications/WEO Keane, M. P., & Wasi, N. (2016). Labour supply: the roles of human capital and the extensive margin. The Economic Journal, 126(592), 578-617. Keane, M. P. (2021). Recent Research on Labor Supply: Implications for Tax and Transfer Policy. Labour Economics, 102026 Kneip, A., Merz, M., & Storjohann, L. (2020). Aggregation and labor supply elasticities. Journal of the European Economic Association, 18(5), 2315-2358. Lama, Ruy. (2011). “Accounting for Output Drops in Latin America”. Review of Economic Dynamics, Vol. 14, pp. 295-316. Levine, R. &Renelt, D. (1992).”A Sensitivity Analysis of Cross-Country Growth Regressions”. American Economic Review, 82(4), 942-963. Lozano, I. & Arias, F. (2018). “Curvas Laffer de la Tributación en Colombia”. Borradores de Economía, 1045. Lucas, R. E. (1988). “On the mechanics of economic development”. Journal of Monetary Economics, 22(1), 3-42. MaCurdy, T. E. (1981). “An empirical model of labor supply in a life-cycle setting”. Journal of Political Economy, 89(6), 1059–85. Mendoza, E. G., Razin, A., &Tesar, L. L. (1994). “Effective tax rates in macroeconomics: Cross-country estimates of tax rates on factor incomes and consumption”. Journal of Monetary Economics, 34(3), 297–323. Ministry of Economy of Argentina (2019). “Economic Data”. Retrieved from: https://www. minhacienda.gob.ar/datos/ Miyamotoy, W. & Nguyen, T. L.& Sergeyev, D. (2016). “Government Spending Multipliers under the Zero Lower Bound: Evidence from Japan,” CEPR Discussion Papers. National Administrative Department of Statistics (DANE) (2019). “Annual National Accounts”. Retrieved from: https://www.dane.gov.co/index.php/estadisticas-por-tema/cuentas-na- cionales/cuentas-nacionales-anuales National Institute of Statistics and Censuses (INDEC) (2019). “Integrated Economic Accounts”. Retrieved from: https://www.indec.gob.ar/indec/web/Nivel3-Tema-3-9 National Institute of Statistics and Geography (INEGI) (2019). “GDP and National Accounts”. Retrieved from: https://www.inegi.org.mx/datos/ National Institute of Statistics and Informatics (INEI) (2019). “Economic Statistics”. Retrieved from: https://www.inei.gob.pe/estadisticas/indice-tematico/economia/ National Tax and Customs Direction (DIAN) (2019). “About the reform”. Retrieved from: https://www.dian.gov.co/impuestos/reformatributaria/aspectosdelareforma/Paginas/ gravamentofinanciero.aspx Nutahara, K. (2015). “Laffer curves in Japan”. Journal of the Japanese and International Economies, 36, 56-72. How do the Tax Burden and the Fiscal Space in Latin America look like? Evidence through Laffer Curves Lozano and Arias 20/26 Observatory of Employment and Business Dynamics (2019). “Statistics and regional indicators”. Retrieved from: http://www.trabajo.gob.ar/estadisticas/oede/estadisticasregionales.asp Organization for Economic Co-operation and Development (OECD) (2019). “OECD Data”. Retrieved from: https://data.oecd.org/ Park, C. (2020). Consumption, reservation wages, and aggregate labor supply. Review of Economic Dynamics, 37, 54-80. Prescott, E. C. (2004). “Why do Americans work so much more than Europeans?”. Quarterly Review, 28(1), 2–13. Ragan, K. S. (2006). “Taxes, Transfers and Time Use: Fiscal Policy in a Household Production Model”. Trabandt, M. &Uhlig, H. (2011). “The Laffer Curve Revisited”. Journal of Monetary Economics, 58(4), 305-327. Trabandt, M. &Uhlig, H. (2012). “How Do Laffer Curves Differ Across Countries?”. National Bureau of Economic Research, Working Paper 17862. Uzawa, H. (1965). “Optimal technical change in an aggregative model of economic growth”. International Economic Review, 6(1), 18-31. Wanniski, J. (1978). “Taxes, revenues and the Laffer curve”. Public Interest, 50, 3-16. Whalen, C., & Reichling, F. (2017). Estimates of the Frisch elasticity of labor supply: a review. Eastern Economic Journal, 43(1), 37-42. Appendix 1. Effective Tax Rates on the Labor, Capital, and Consumption Technical Details In this section we provide details on the calculation of effective tax rates on labor, capital income and consumption. The main source of information is the National Statistics Bureau for each country, specifically the Integrated Economic Accounts. We also use complementary information from the OECD - mainly for the case of Argentina - and Penn World Tables databases. Table A1.1 includes the definitions for each effective tax rate, while Table A1.2 describes all the variables used in their calculations. Table A1.1. Effective Tax Rates, based on Mendoza et al. (1994). Effective Tax Formula Consumption Personal Income Labor Income Capital Income Source: Authors’ calculations. Latin american economic review (2021) 30:9 21/26 Table A1.2. Variable description. Variable Description INN Indirect taxes on production of goods and services. ICIH Income, profit, and capital gains taxes of individuals. ICIE Income, profit, and capital gains taxes of corporations. CST Social security contributions, total. CSE Social security contributions made by employers. ICNOM Payroll Taxes. ITF Financial transactions tax. CFinal household expenditures. GC Final government expenditures. GW Salaries paid by the government. W Wage income received by households. ENE Gross operating surplus minus consumption of fixed capital. IMKMixed income attributable to capital. IML Mixed income attributable to labor. RNPH Net property income, households. ENEH Net operating surplus, households. Source: Authors’ calculations. The sources of all the data used are listed by country: • Argentina: the main sources of data are the OECD and the National Institute of Statistics and Censuses (INDEC by its Spanish acronym). The tax related data was retrieved from the OECD, and the data such as consumption and gross operating surplus was retrieved from the Integrated Economic Accounts (IEA) found on the INDEC’s website. Only the wages paid by the government (GW) and the tax on payroll (ICNOM) were found elsewhere. GW was taken from the Ministry of Economy, and the ICNOM was estimated with information from the provincial tax administrations and the Observatory of Employment and Business Dynamics. • Brazil: most of the data was retrieved from the IEA found on the Brazilian Institute of Geography and Statistics (IBGE by its Portuguese acronym). Only the ICNOM and the tax on financial transactions (ITF) were taken from other databases. The ICNOM was retrieved from the OECD and the ITF from the Inter-American Center of Tax Administrations (CIAT by its Spanish acronym). • Chile: the data was obtained from Chile’s Central Bank; however, it is important to note that the ICNOM and ITF were not found, hence they are not included in the calculation of Chile’s effective tax rates. • Colombia: the main data source was the National Administrative Department of Statistics (DANE by its Spanish acronym). Only two variables were found elsewhere. The ICNOM was estimated from contributions made to family compensation funds and the ITF was retrieved from the National Tax and Customs Direction (DIAN by its Spanish acronym). How do the Tax Burden and the Fiscal Space in Latin America look like? Evidence through Laffer Curves Lozano and Arias 22/26 • Mexico: most of the data was retrieved from the IEA found on the National Institute of Statistics and Geography (INEGI by its Spanish acronym). Only the ICNOM and the ITF were taken from other databases. The ICNOM was retrieved from the OECD and the ITF from the CIAT. • Peru: the main data source was the National Institute of Statistics and Informatics (INEI by its Spanish acronym). Only the ICNOM and the ITF were taken from other institutions. The ICNOM was retrieved from the OECD and the ITF from the CIAT. Appendix 2. Laffer Curves for Labor Income Taxes in Latin America: Exogenous and Endogenous Growth Models Argentina Brazil Chile Colombia Mexico Peru Source: Authors’ calculations. Latin american economic review (2021) 30:9 23/26 Appendix 3. Laffer Curves for Capital Gains in Latin America: Exogenous and Endogenous Growth Models Argentina Brazil Chile Colombia Mexico Peru Source: Authors’ calculations. How do the Tax Burden and the Fiscal Space in Latin America look like? Evidence through Laffer Curves Lozano and Arias 24/26 Appendix 4. Laffer Curves in LAC for Labor and Capital Income Using Statutory Tax Rates In this appendix we perform a Laffer Curve analysis based on the same parameters as in Table 1 but employing statutory tax rates instead of effective tax rates. The source of the information is the CIAT and the OECD. Table A4.1 describes the rates we use. For labor we work with the maximum marginal tax rate applied to personal income, while for capital we employ the country’s legal general tax rate plus any surcharge incorporated in its tax codes. Table A4.1. Set of Statutory Tax Rates for Capital and Labor Income - 2017 Country Statutory Tax Rate on Labor Income* Statutory Tax Rate on Capital Income Argentina 0.35 0.35 Brazil 0.275 0.34 Chile 0.35 0.25 Colombia 0.33 0.40 Mexico 0.35 0.30 Peru 0.30 0.28 LAC 0.306 0.318 * For each country, we use the maximum marginal tax rate applied to personal income. The Laffer curve estimates for taxes on labor (Panel A) and capital income (Panel B) for the major LAC are presented in Graph A4.1. Each curve shows that tax rates that maximizes revenues using statutory rates are lower (Graph A4.1. vs Graph 5). Therefore, the fiscal space is smaller than in the effective tax rates case and this occurs in both human capital versions of our modelling. The graphs suggest that the statutory rate that maximizes government revenues coming from labor would be around 63% contrasting with the 69% we got on the effective rates estimates (Graph 5) and the rate that maximizes from capital income would be 44% comparing to the 51% with effective rates. As expected, the fiscal space in terms of additional revenues for the government also will be much lower using statutory rates. Government revenues would increase roughly 15%, on the hypothetical scenario that the rate would be increased to its maximizing level, for the labor income tax (contrasting with the 31.8% of the effective rates estimates) and the revenues would rise only by 0.96% for the capital income tax (comparing with the 9.14% on the initial case). Across countries the results are consistent with the trend above. Nonetheless, one exception is Brazil, where its labor statutory tax rate happens to be higher than its effective tax rate. Thus, there is more fiscal space using the statutory rates. This result is in line with the main idea pertaining to the direct relationship between the size of the tax and the distortions it will bring into the economy.