Tax reform in Japan: Is it welfare-enhancing? *

Transcription

1 Tax reform in Japan: Is it welfare-enhancing? * Yoonseok Choi ** Korea University Hideaki Hirata *** Hosei University Sunghyun Kim **** Sungkyunkwan University Abstract The ongoing tax reform in Japan, in particular a corporate income tax cut financed by an increase in VAT, is expected to bring positive effects on investment and output. However, the overall effects on government fiscal balance and welfare can be ambiguous depending on dynamic responses of macroeconomic variables to changes in tax rates. This paper aims to provide quantitative forecasts of Japanese tax reform on welfare and fiscal balance using a small open two-sector dynamic general equilibrium model calibrated to the Japanese economy. Simulation results show that under unrestricted international borrowing and lending and no habit, a corporate income tax cut of 2.5% financed by an increase in VAT improves welfare by 0.18%. However, positive effects of corporate income tax cut disappear when international borrowing becomes limited or consumers show habit formation. JEL Classification: E6 Key Words: corporate income tax, Japan, dynamic scoring, revenue neutral, tax reform. * This work is supported by the National Research Foundation of Korea Grant funded by the Korean Government (NRF-2014S1A5B ) and Japan s Grant-in-Aid for Scientific Research (C, ). ** Department of Economics, Korea University, Seoul, Korea. yoonchoi3@korea.edu. *** Department of Business Administration, Hosei University. h-hirata@hosei.ac.jp. **** Corresponding Author. Department of Economics, Sungkyunkwan University, Seoul, Korea. Tel: , shenrykim@skku.edu. 1

2 1. Introduction Japan has been implementing a series of tax reform starting from Two main objectives in Japanese tax reform are improving growth and restoring fiscal balance especially with impending explosion in social security spending due to aging population. The main direction of tax reform is to lower corporate income tax that is expected to boost corporate profitability, promote inward investments from the rest of the world, thereby bolstering economic growth. The Japanese government plans to finance the lost tax revenue from corporate income tax cut by increasing consumption tax (VAT). The Abe administration has committed to lower the effective corporate tax rates from 34.6% to below 30% and to raise VAT up to 10% from 5%. 1 How much positive effect can this tax reform generate on investment and output? How would the fiscal revenue and welfare respond over time? These are important questions to answer in order to successfully implement the tax reform program in Japan. This paper aims to provide quantitative effects of potential Japanese tax reform on welfare and fiscal budget balance using a small open economy two-sector dynamic general equilibrium model. We use a dynamic scoring technique to measure the overall effects on fiscal balance under various scenarios of tax reform. The baseline model is similar to the one used in Choi and Kim (2014) where two main improvements were made compared to the previous literature: open economy and two-sector (tradable and nontradable) model. Since Japanese economy is open to international borrowing and lending with a large trade sector, it is important to incorporate these features in the model. In general, tax reform involving a corporate income tax cut financed by an increase in consumption tax is welfare enhancing (e.g. Kim and Kose, 2004; Mendoza and Tesar, 1998). 2 An increase in consumption tax rate is likely to be less distortionary than other sources of 1 In 1989, the Japanese government introduced 3% VAT with a significant reduction in corporate tax and income tax. In 1997, VAT was raised to 5% with an advanced implementation of a tax cut for three subsequent years. In April 2014, VAT was raised to 8% and a fiscal stimulus of 5 trillion Yen was implemented to mitigate a short-term negative impact of the VAT hike. It was expected that the Japanese government would increase VAT to 10% in September 2015, but the Abe administration postponed it to April Heer and Trede (2003) also constructed a dynamic general equilibrium model for Germany and found that a consumption tax which replaces income tax generates only a small impact on labor income distribution, but positive effects on employment and long-run welfare level. 2

3 fiscal revenue such as income taxes (IMF, 2011&2013). 3 Some studies in dynamic scoring show that corporate income tax can be self-financing especially in open economies and in the long run (e.g. Choi and Kim, 2014). 4 Therefore, the overall effects of a corporate income tax cut on fiscal balance may not be as bad as one imagines. Exact quantitative forecasts depend on the structure of model and parameter values used for model simulation. Dynamic scoring, a method employed in this paper, is widely used to measure dynamic revenue effects of tax reform in recent studies. Using a simple neoclassical growth model, Mankiw and Weinzierl (2006) find that the long-run feedback effect on revenue for capital income taxes is 50 percent. Trabant and Uhlig (2011) find similar results for the US and larger effects for Europe. For Japan, Nutahara (2013) reports that the capital tax rate is close to the top of the Laffer curve, i.e. feedback effects are close to 100 percent. Hiraga (2011) even reports that a 5 percentage point corporate income tax cut in Japan will increase tax revenue by around 50 percent in the long run. Yet, other models yield (much) smaller effects, depending on their structure and parameter calibration. Ganelli and Tervala (2014) employ a New-Keynesian two-country model to assess the difference between short and long-run effects on the budget. They report feedback effects of income taxes in the order of 22 percent in the long run, while short-run effects might in fact be negative. Some papers have analyzed the effects of a corporate income tax cut on investment and welfare in the case of Japan. 5 DeMooij and Saito (2014) argue that a 1% reduction in 3 Unlike capital income tax, consumption tax does not distort household savings decisions, investment decisions or trade, and a part of the burden of an increase in tax rate falls on spending financed by past savings, which is completely distortion free since past decisions cannot be changed. 4 Dynamic scoring calculates revenue effects of a proposed tax policy using dynamic macroeconomic models, in which a change in tax rates generates feedback to tax revenues through changes in tax base over time. For example, a tax cut in capital income (or other distortionary taxes) tends to reduce tax revenue on impact but it increases tax base (and tax revenue) over time because a tax reduction in capital income enhances economic activities such as investment. The first effect of a tax cut is called static scoring, while the second effect (related to an increase in tax base) is called the feedback (dynamic) effect. Compared to dynamic scoring, static scoring tends to overestimate the fiscal revenue loss from a tax cut as it disregards changes in tax base. 5 Several papers have analyzed the fiscal sustainability issue of tax reform, particularly an increase in consumption tax. Most existing literature suggests that consumption tax rate should be higher than 20% in order to halt the increase in government-debt-to-output ratio. Braun and Joines (2014) suggest that consumption tax rate should increase to 33% in 2017 to maintain fiscal balance assuming that consumption tax rate reaches 10% in Sakuragawa and Hosono (2011) claim that consumption tax rate should increase by 16% points, i.e., from 5% to 21% in 2031, to raise primary surplus-to-gdp ratio by 6.5% points. Hoshi and Ito (2014) and Fukao (2012) find the government needs to increase consumption tax up to nearly 25% for stabilizing the governmentdebt-to-output ratio. Hansen and Imrohoroglu (2013) argue that the revenue required to finance the projected 3

4 corporate income tax rate can boost investment by 0.4%. Employing a Bayesian DSGE model, Hasumi (2014) studies the case of a corporate tax reduction (matching 1% GDP scale) replaced by an increase in consumption tax and finds that output increases by 1.1% within two years and that consumption drops on impact but rises later on. This paper analyzes the welfare effects of a 2.5% cut in corporate tax rate financed by an increase in lump-sum tax or consumption tax. In the benchmark case with unrestricted international borrowing and lending and no habit, such corporate income tax cut financed by consumption tax increases welfare by 0.18%. As restrictions on international borrowing and lending or the intensity of habit persistence increase, the amount of welfare gains decreases. For example, in the closed economy model or with strong habit persistence, the tax reform yields welfare loss. Next, we focus on the effects of tax reform on fiscal balance using dynamic scoring. Capital income tax cut of 2.5% in both sectors financed by lump-sum tax (government debt) generates revenue loss of 0.33% in the baseline economy. 6 When financed by an increase in consumption tax, such tax cut generates revenue gain of 1.27%. The fiscal revenue calculated by static scoring tends to overestimate revenue loss relative to dynamic scoring because static scoring does not take account of change in tax base over time, which affects fiscal revenue in the future. There are several other important points that need to be considered to study Japanese tax reform. First, a rise in consumption tax is expected to help improve fiscal revenue in the midst of population aging. IMF (2010) documents Aging implies dissaving by households (spending exceeding income), making the VAT base more robust than that of taxes on labor income, which will grow more slowly than spending as the population ages. Second, the currently proposed tax reform scenario may be able to improve distortions in the current Japanese fiscal revenue structure: too low consumption tax revenue with too high corporate tax revenue (See figures 1 and 2). Finally, an increase in consumption tax has negative side effects on income distribution as consumption tax is regressive. Consumption tax has a broader base than income tax because it also falls on spending financed by social transfers. increase in government expenditure and to stabilize Japanese government debt is in the order of 30-40% of aggregate consumption per year. 6 Note that lump-sum tax does not count as fiscal revenue in this paper as it is equivalent to government debt. 4

5 The paper is organized as follows. Section 2 lays out a small open two-sector dynamic general equilibrium model. Section 3 calibrates the model with explanations on deep parameters. Section 4 provides the main results from various exercises and presents results from sensitivity check. Section 5 concludes. 2. The Model The model consists of two sectors, tradable and nontradable sectors. Households supply labor and capital to firms. Their labor income and capital income are subject to taxes and the households also pay a tax on their consumption. Firms use two factors, labor and capital, to produce two final goods, tradable and nontradable goods. The model allows for international borrowing and lending in the form of one-period risk-free bonds. The government finances an exogenous stream of expenditures through domestic taxes. A representative household solves t, xt, nt, where t (1) Max β U ( c h h ) t= 0 U t θ ( c ηc ) ( h h ) 1 1+ ξ t t 1 xt + nt = ϕ 1 θ 1+ ξ, subject to the budget constraint x (2) ( 1 ) 1 ( 1 )( ) ( 1 ) + τ pc + i + p i + B+ = τ w h + p w h + τ r τ δ + k x ct t t xt nt nt t ht xt xt nt nt nt kt xt kt x xt n n ( 1 τkt ) rnt τktδ n pntknt pnttt RB t t, where,,,, (,,,, ) w r h i k w r h i k are wage rate, rental rate, hours worked, xt xt xt xt xt nt nt nt nt nt investment and capital for the tradable (nontradable) sectors. The parameters θ, η, φ and ξ in the utility function denote the inverse of elasticity of intertemporal substitution (EIS), the degree of external habit persistence, the disutility parameter for labor and the Frisch elasticity 5

6 of labor supply, respectively. 7 β is the discount factor and ( ) δ δ is depreciation rate for the x n tradable (nontradable) sector. The price of composite consumption good c is p. t t Bt is international bonds purchased in period t maturing in t+1. interest rate on bonds. n kt Rt is the exogenously given gross Tt is lump-sum transfers from the government, and τ is tax rate ( τ ht = labor income tax, τ = tax on capital income from the nontradable sector, τ = tax on capital income from the tradable sector, and τ ct = consumption tax). Investment tax credit is incorporated in the budget constraint. All the prices are normalized in terms of tradable good ( p xt ), which means that pnt is the price of the nontradable good in terms of the price of the tradable good (reciprocal of the real exchange rate). Note that corporate income tax and capital income tax are equivalent in this model where households implicitly hold 100% of firm s shares. Therefore, we use capital income tax instead of corporate income tax in this paper. Capital in each sector is subject to adjustment costs as in Baxter and Crucini (1993) x kt (3) ( δ ) (4) ( δ ) i xt kx, t+ 1 = 1 x kxt + φ kxt, kxt i nt knt, + 1 = 1 n knt+ φ knt, knt where () φ is the adjustment cost function with the following properties: φ() φ () and φ ''() < 0. > 0, ' > 0 A composite consumption good consists of two goods, tradable ( cxt ) and nontradable ( cnt ) goods. The optimal consumption level of the tradable and nontradable goods can be derived from the expenditure minimization problem 7 This additively separable utility functional form is frequently used in the open macro literature (e.g., Erceg et al. 2005; Galí et al. 2005). The external habit formation is sometimes called catching-up-with-the-jones preference, which is widely used in macro literature. Iiboshi et al. (2006) and Smets Wouters (2007), for example, provide the estimates of the degree of external habit by estimating the dynamic stochastic general equilibrium (DSGE) models using the Japanese and U.S. data, respectively. 6

7 (5) Min p c c p c, s.t. t t xt nt nt 1 1 γ 1 γ 1 t = γ x xt + n nt c b c b c and bx + bn = 1, where b x and b n are share parameters representing a relative weight to consumption of each good and γ is a parameter related to the price elasticity for two goods (i.e., inverse of γ is the price elasticity of demand for tradable and nontradable goods). Firms in the tradable and nontradable sectors produce output using labor and capital based on the typical Cobb-Douglas technology with zero profit conditions (6) y = w h + r k, y = A k h μ 1 xt xt xt xt xt xt xt xt xt (7) y w h r k y A k h α 1 nt = nt nt + nt nt, nt = α nt nt nt, where y ( y ) denote output of the tradable (nontradable) sector and μ ( α ) indicates the xt nt capital share of the tradable (nontradable) sector. Axt and A nt are defined as productivity in production functions and are assumed to be constant at one in this deterministic model. The government s budget constraint is given by x n (8) τ p c + τ ( w h + p w h ) + τ ( r δ ) k + τ ( r δ ) p k = p G + p T, ct t t ht xt xt nt nt nt kt xt x xt kt nt n nt nt nt nt nt t where Gnt is exogenous government spending on the nontradable good. We assume that government spending and lump-sum transfers are in the nontraded sector only. Combining the household s budget constraint with the government s budget constraint yields the following aggregate budget constraints (9) y = c + i + G, nt nt nt nt (10) y + RB = c + i + B. + 1 xt t t xt xt t 7

8 3. Calibration We calibrate the model parameters based on the data of Japan and adopt some parameter values from the previous literature. Table 1 reports the parameters and steady state values of the benchmark model. The discount factor β is set at 0.95 to match the annual steady state world real interest rate of about 5%. 8 The inverse of elasticity of intertemporal substitution (EIS) θ is equal to 1.9, implying a low EIS that is commonly observed in DSGE literature. The parameter for habit persistence η is set at 0, 0.2 and 0.5 for different types of experiments. These values fall within commonly used values in the literature (e.g., Christiano et al. 2005; Smets and Wouters 2007). The inverse of Frisch elasticity of labor supply ξ is set to equal 2.1, suggesting a low Frisch elasticity that is close to that used in Smets and Wouters (2007). 9 The labor disutility parameter ϕ is set to assure that the steady state value of hours worked (i.e., sum of hours worked in the tradable and nontradable sectors) is one-third. Share parameters bx and b n in the CES form of consumption function are set to match the actual consumption shares in the Japanese data during The data show that the consumption share of the tradable and nontradable goods is 43% and 57%, respectively. 10 The parameter γ is set at 2 to make the elasticity of substitution between tradable and nontradable goods in Japan equal to 0.5 following Batini et al. (2005). 11 We set the depreciation rate at 8% for both production sectors following Hayashi and Prescott (2002), which is close to a commonly used value in the literature. The elasticity of the marginal adjustment cost functionφ of the tradable and nontradable sectors is set at 25 to match the relative volatility of investment. 12 The capital share in the tradable sector μ is set at 0.44 and 8 The U.S. 3-month Treasury bill rate is used as a proxy for the world interest rate. We calculate the longrun annualized average T-bill rate from 1970 to The EIS and the Frisch elasticity of labor supply are taken from Iiboshi et al. (2006) who employ the Bayesian estimation method to estimate a DSGE model using the Japanese data. 10 The consumption share data are from ESRI, Annual Reports on National Accounts. 11 Some earlier studies show that the elasticity of substitution between tradable and nontradable goods is less than unity. For example, Mendoza (1995) and Stockman and Tesar (1995) use 0.74 and 0.44, respectively, which correspond to the parameter of about 1.35 and 2.27, respectively. The value used in this paper is between these two values. 12 Iiboshi et al. (2006) estimate a quadratic capital adjustment cost function using the Japanese data and obtains 0.04 as a posterior mean of the inverse of capital adjustment cost function. 8

9 that in the nontradable sector α is set at 0.35 to match the long-run average labor income share in the Japanese data. 13 Mendoza et al. (1994) calculate effective tax rates for G-7 countries by dividing actual tax payments by corresponding national accounts. In this paper, we follow this method and calculate the aggregate effective tax rates of Japan. The data show that the computed tax revenues from consumption, labor income and capital income taxes are 16.9%, 55.7% and 27.5%, respectively. 14 The effective tax rates in Japan are set at 8.16%, 28.78% and 40.54% for consumption, labor and capital income tax, respectively, in order to match the tax revenue shares data. We use these values for the steady state tax rates ( τc, τh and τ k ) in the model economy. The government expenditure on the nontradable sector Gn over yn is set at 21.94% to match the government expenditure over GDP ratio at 17.39%, which allows the government budget to be balanced under the steady state value. The initial asset holding position (which is a free parameter) is set to zero and p xt is set to one. We solve the model using a double shooting algorithm employed by Mendoza and Tesar (1998), Gorodnichenko et al. (2012), and Kim and Kose (2014) Main Results Since the results are sensitive to two main assumptions in the model (degree of capital mobility and intensity of habit formation), we experiment with different values for these two parameters. The benchmark is when there are no restrictions in international borrowing and lending and no habit formation is present. Then, we experiment with low and high degree of restrictions in international borrowing and lending, and positive habit formation. 13 Freeman (2008) classifies the tradable and nontradable sectors using the OECD dataset, which is based on the International Standard Industry Classification (ISIC Rev.3) and employs the "industry" ("market services") in the OECD dataset as the tradable (nontradable) sector. Following Freeman (2008), we compute the long-run average of labor income share from 1970 to Data are taken from the OECD dataset from 2000 to For detailed shooting algorithm, see Mendoza and Tesar (1998). 9

10 4.1. Measuring consumption tax rate to obtain the balanced budget In this section, we measure how much a consumption tax rate should increase in order to obtain a balanced budget over the next five years. We set the initial budget balance at the deficit of 6.29% of GDP, which is equal to the average rate between 2000 and The initial budget deficit is financed by lump-sum tax (government debt) which will be replaced by consumption tax in this exercise, while holding other tax rates intact. This exercise would give us a rough idea of how much consumption tax rate should be in order to improve fiscal situation in Japan. Table 3 reports the results of the amount of consumption tax hike under different financial market structure and habit persistence. The baseline case of unrestricted international borrowing and lending and no habit shows that a consumption tax rate should increase from 8.16% to 18.35% in order to obtain the balanced budget over the next five years. As the habit persistence becomes stronger, the amount of consumption tax rate hike necessary to achieve the balanced budget decreases. For example, compared to the new consumption tax rate of 18.35% in the benchmark case, when the degree of habit persistence is 0.5, a required consumption tax rate for the balanced budget is lower at 18.04%. With a stronger habit behavior, consumption becomes less responsive to changes in tax rates and therefore consumption tax base decreases less when the consumption tax rate increases. Therefore, a smaller increase in consumption tax rate should be sufficient to obtain the balanced budget compared to the case of no habit. An increase in restrictions on borrowing and lending leads to a higher consumption tax rate required for the balanced budget. In an extreme case, for instance, the closed economy requires a new consumption tax rate to be at 18.55%. Overall, the results imply that the consumption tax rate should increase by at least 10% in order to obtain the balanced budget over the next five years and this number is quite consistent under different assumptions on international financial markets or consumption habit persistence. 16 The data are taken from OECD dataset. 10

11 4.2. Capital income tax cut in both sectors We conduct two policy experiments a one-time permanent cut in capital income tax financed by lump-sum tax or consumption tax. We examine how macroeconomic variables, in particular tax revenue and budget balance, change over time using dynamic scoring. We also measure welfare effects in the short-run and long-run Financed by lump-sum tax In this section, we consider a 2.5% tax cut in capital income (from 40.54% to 38.04%) financed by lump-sum tax. Table 4 reports welfare effects under various financial market structures and habit formation. Welfare gains are measured by percentage changes in certainty equivalent consumption, which is frequently used in macro public finance literature. Overall welfare gains are measured by conditional welfare changes which are calculated by taking a discounted sum of periodic utility (measured by certainty-equivalent consumption) over time. Long-run welfare gains are measured by changes in steady state welfare before and after tax reform. The difference between overall welfare gains and long-run gains are defined as transitional gains. We consider four financial market structures: unrestricted borrowing, restricted borrowing, very restricted borrowing and no borrowing. We use different values for bond holding costs for each financial market structure in the model. For the restricted borrowing case (very restricted borrowing), we choose bond holding costs so that the bond holding over GDP ratio in the new steady state is at 75% (50%) of that from the unrestricted borrowing case (around 7.6%). The simulation results show that there exist sizeable long-run welfare gains in all cases. In the long run, capital income tax cut financed by lump-sum tax (government debt) is always beneficial. However, transitional welfare gains are negative in all cases, which indicates that capital income tax cut reduces consumption and/or increase labor supply over transitional period, which lowers welfare. Capital income tax cut increases investment but the increase in investment is accompanied by a decrease in consumption due to substitution effects. The size of transitional welfare loss depends on financial market structure and intensity of habit persistence. As the bond holding costs and habit persistence increase, transitional welfare loss becomes larger and therefore overall welfare gains decrease. Under higher bond holding costs, 11

12 an increase in investment requires a sacrifice of consumption instead of international borrowing, which lowers welfare. Strong habit persistence implies that consumption does not move to an optimal level when tax rate changes and therefore welfare gains decrease. In terms of overall welfare gains, open economy model with unrestricted borrowing and no habit formation yields substantial welfare gains of 0.23%. Overall welfare gains stay positive as long as there is no habit even under the closed economy. However, with higher habit persistence and bond holding costs, welfare gains become negative. The top panel of Table 7 shows the long-run steady state changes in tax base and tax revenue in both open and closed economies in the case of no habit. The reported numbers are percentage changes between the initial and new steady states. The results exhibit significant feedback effects as all three tax bases increase in the long run consumption tax base by 0.73%, labor income tax base by 1.98% and capital income tax base by 0.12%, respectively in the open economy. The total tax revenue under dynamic scoring decreases by 0.33% from the initial steady state. This value is much smaller than that of static scoring, 1.60%. It implies that static scoring tends to overestimate revenue loss due to an absence of feedback effects captured by changes in tax base over time. Note that the lump-sum tax revenue is not included in total tax revenue. Lump-sum tax financing is equivalent to bond financing by government. That is, government borrows from households by using one period government bond (this is different from international bond that households trade). Therefore, lump-sum tax revenue is considered as government borrowing, not part of total tax revenue Financed by consumption tax We compare how the previous results change when the government raises a consumption tax instead of the lump-sum tax to finance lost tax revenue from the capital income tax cut. We calculate the necessary amount of increase in consumption tax rate by using a double shooting algorithm. A 2.5 percent tax cut in capital income can be compensated fully by an increase in consumption tax from 8.16% to 8.99%. Note that the necessary amount of increase in the consumption tax rate is calculated to match government s intertemporal budget constraint for 10 years, not period-by-period budget balance. Since we 17 See Mendoza and Tesar (1998) for a detailed explanation. 12

13 only consider a time-invariant one-time change in tax rate, lump-sum taxes or transfers are used to match the budget balance in every period. Table 5 shows the result of a 2.5% tax cut in capital income financed by a consumption tax. As in the case of lump-sum tax financing, long-run welfare gains are positive in all cases. Overall welfare gains depend on transitional change: as bond holding costs and habit persistence increase, overall welfare gains decrease due to larger transitional losses. Open economy model with unrestricted borrowing and no habit yields the largest overall welfare gain, 0.18%. Compared to the case of lump-sum tax financing, welfare gains of capital income tax cut with consumption tax financing is smaller in all cases. This is because lumpsum tax is the least distorting taxing method. In fact, lump-sum tax does not generate any distortions in optimality conditions. The bottom panel of Table 7 reports the effects on tax revenue in the case of consumption tax financing. As in the case of lump-sum tax financing, dynamic scoring leads to substantial feedback effects: consumption and labor tax bases increase by 0.55% and 1.76%, respectively in the open economy. The total tax revenue under dynamic scoring increases by 1.27% from the initial steady state, which is much larger than that of static scoring, We can observe a noticeable difference between two financing schemes in tax revenue and budget balance. In the case of lump-sum tax financing, a 2.5% tax cut in capital income generates only small revenue recoveries because lump-sum tax revenue is not included in total tax revenue as it is considered as government debt in this paper Capital income tax cut in one sector only One of the advantages of using a multi-sector model is that we can analyze the effects of capital income tax cut in one sector only. Table 6 reports the welfare effects from a 2.5% capital income tax cut in one sector, holding the tax rate in the other sector constant. One salient feature of the result is that the effects of a tax cut in the tradable sector are much better than those in the non-tradable sector in terms of overall welfare. This result holds regardless of financial market structure or the degree of habit formation. In open economy model with no habit, for instance, overall welfare gains of a tax cut in the tradable sector are 0.25%, while the tax cut in the non-tradable sector generates welfare loss of 0.02%. As the bond 13

14 holding costs and habit persistence increase, transitional welfare loss becomes larger and therefore overall welfare gains decrease. The middle panel of Table 7 shows the effects on tax revenue from a capital income tax cut in one sector only. Overall, the positive revenue effects of a tax cut in the tradable sector is much larger than those in the non-tradable sector in the open economy. For example, a tax cut in the tradable sector increase consumption, labor and capital tax bases by 1.16%, 1.86% and 1.08%, respectively. However, a tax cut in the non-tradable sector show mixed results: labor income tax base increase by 0.11%, but consumption and capital tax bases decrease by 0.43% and 0.95%, respectively. All of these results translate into fiscal revenue. Tax revenue increases by 0.84% from a tax cut in the tradable sector, while it decreases by 1.15% when the tax cut is implemented in the nontradable sector. Since the tradable sector is more capital intensive than the nontradable sector in the Japanese economy (capital income share is 0.44 vs in calibration), a capital income tax cut is more beneficial when it is applied to the tradable sector. Main reason behind this asymmetric result comes from sectoral shift of resources. 5. Conclusion We examine welfare effects and dynamic revenue effects of a permanent tax cut in capital income using a small open two-sector dynamic general equilibrium model. We evaluate the consequences on tax revenue using dynamic scoring. All the results show that the feedback effect is not negligible and static scoring overestimates the revenue loss from a tax cut. Several findings emerge. First, capital income tax cut in the open economy model with unrestricted borrowing and no habit yields positive welfare gains, whether the tax cut is financed by lump-sum tax or consumption tax. More restrictions on international borrowing or larger habit persistence reduce overall welfare gains and in many cases, yield welfare loss. Second, positive feedback effects of a tax cut are much larger in the open economy model, compared to closed economy case. Households do not need to sacrifice consumption in order to take advantage of a capital income tax cut and produce more as they can rely on international borrowing. Third, sectoral responses to a capital income tax cut exhibit a stark 14

15 contrast in tradable and nontradable sectors. Since the tradable sector is more capital intensive than the nontradable sector in this model, there is a significant resource shift from the nontradable to tradable sector when there is a capital income tax cut. This result provides policy implication that if the government can implement a selective capital income tax cut in different sectors, a tax cut in the tradable sector generates better revenue effects than the case of a tax cut in the nontradable sector. 15

16 References Auerbach, A.J., Dynamic scoring: an introduction to the issues. American Economic Review 95, Batini N., P. N'Diaye and A. Rebucci, The domestic and global impact of Japan s policies for growth. IMF working paper, WP/05/209. Baxter, M. and M.J. Crucini, Explaining saving-investment correlations. American Economic Review 83, Braun, R.A. and D.H. Joines, The implications of a graying japan for government policy. Working Paper , Federal Reserve Bank of Atlanta. Carey, D. and H. Tchilinguirian, Average effective tax rates on capital, labour and consumption. OECD Working Paper, No Choi, Y.S. and S.H. Kim, Dynamic scoring of tax reform in small open economies. Working paper. Christiano, L. J., M. Eichenbaum and C.L. Evans, Nominal rigidities and the dynamic effects of a shock to monetary policy. Journal of Political Economy 113, Cooley, F.T. and G.D. Hansen, Tax distortions in a neoclassical monetary economy. Journal of Economic Theory 58, Erceg, C.J., L. Guerrieri and C. Gust, Expansionary fiscal shocks and the US trade deficit. International Finance 8, Freeman, R., Labour productivity indicators: Comparison of two OECD databases productivity differentials & the Balassa-Samuelson effect. OECD Statistics Directorate, Division of Structural Economic Statistics. Fukao, M., 2012.The sustainability of budget deficit in Japan (Japanese). Discussion Papers 12018, Research Institute of Economy, Trade and Industry (RIETI). Galí, J. and T. Monacelli, Monetary policy and exchange rate volatility in a small open economy. Review of Economic Studies 72, Gorodnichenko, Y., E.G. Mendoza and L.L. Tesar, The Finnish great depression: from Russia with love. American Economic Review 102, Greenwood, J. and G.W. Huffman, Tax analysis in a real-business-cycle model: on measuring Harberger triangles and Okun gaps. Journal of Monetary Economics 27,

17 Heer, B. and M. Trede, Efficiency and distribution effects of a revenue-neutral income tax reform. Journal of Macroeconomics 25, Hoshi, T. and T. Ito, Defying gravity: can Japanese sovereign debt continue to increase without a crisis? Economic Policy 77, Iiboshi, H., S.-I. Nishiyama, T. Watanabe, An estimated dynamic stochastic general equilibrium model of the Japanese economy: A Bayesian analysis. Mimeo. Jonsson, K., Real exchange rate and consumption fluctuations following trade liberalization. Sveriges Riksbank Working Paper 187. Kim, S.H. and M.A. Kose, Welfare implications of trade liberalization and fiscal reform: A quantitative experiment. Journal of International Economics 92, Leeper, E.M. and S.C.S. Yang, Dynamic scoring: alternative financing schemes. Journal of Public Economics 92, Mankiw, N.G. and M. Weinzierl, Dynamic scoring: a back-of-the-envelope guide. Journal of Public Economics 90, Mendoza, E.G., The terms of trade, the real exchange rate, and economic fluctuations. International Economic Review 36, Mendoza, E.G. and L.L. Tesar, The international ramifications of tax reforms: supplyside economics in a global economy. American Economic Review 88, Mendoza, E.G. and L.L. Tesar, Why hasn't tax competition triggered a race to the bottom? Some quantitative lessons from the EU. Journal of Monetary Economics 52, Mendoza, E.G., A. Razin and L.L. Tesar, Effective tax rates in macroeconomics: crosscountry estimates of tax rates on factor incomes and consumption. Journal of Monetary Economics 34, Novales, A. and J. Ruiz, Dynamic Laffer curves. Journal of Economic Dynamics and Control 27, Ryo, H., The effect of corporate tax rate reduction: a simulation analysis with a small open economy DSGE model for Japan. Discussion Papers 14040, Research Institute of Economy, Trade and Industry (RIETI). Sakuragawa, M. and K. Hosono, Fiscal sustainability in Japan. Journal of the Japanese and International Economies 25, Smets, F. and R. Wouters, Shocks and frictions in US business cycles: A Bayesian DSGE approach. American Economic Review 97,

18 Stockman, A.C. and L.L. Tesar, Tastes and technology in a two-country model of the business cycle: Explaining international comovements. American Economic Review 85, Strulik, H. and T. Trimborn, Laffer strikes again: dynamic scoring of capital taxes. European Economic Review 56, Trabandt, M. and H. Uhlig, The Laffer curve revisited. Journal of Monetary Economics 58,

19 Figure 1. Value added and sales tax rates % Canada Japan Switzerland Australia Korea Luxembourg NZL Israel Mexico Turkey Chile Germany France Austria Estonia Slovak Slovenia United Kingdom Belgium Czech Netherlands Spain Italy Greece Ireland Poland Portugal Finland Denmark Norway Sweden Iceland Hungary Sources: OECD, Tax Database. Figure 2. Tax revenues by sources % JPN GER FRA GBR USA Personal Income Corporate Income Consumption Tax Others Source: OECD, Revenue Statistics Notes: Each item corresponds to the source of tax revenue in Social Security is excluded. 19

20 Table 1. Deep parameters and key steady state values Parameters Description Benchmark Values Preference β Discount factor (annual) 0.95 θ Inverse of elasticity of intertemporal of 1.90 substitution (EIS) η Degree of external habit persistence 0.80 ξ Frisch elasticity of labor supply 2.10 γ Coefficient of elasticity of substitution b/w tradable and nontradable goods 2.00 b Weight of tradable good 0.43 x b Weight of nontradable good 0.57 n Technology Tradable Sector μ Share of capital income 0.44 δ x Depreciation rate (annual value) 0.08 η x Elasticity of marginal adjustment cost function 25 Nontradable Sector α Share of capital income 0.35 δ n Depreciation rate (annual value) 0.08 η n Elasticity of marginal adjustment cost function 25 Other Steady States g Government expenditure (ratio of GDP) 21.94% n nx Net exports (ratio of GDP) 0 p x Price of tradable good 1.00 Tax Rates τ Consumption tax 8.16% c τ h Labor income tax 28.78% x n τ, τ Capital income tax 40.54% k k 20

21 Table 2. Tax rates in selected countries Average effective tax rates (%) C-tax rate L-tax rate K-tax rate Japan G-7 countries in Mendoza et al. (1994) Average Range OECD countries in Carey and Tchilinguirian (2000) average range Note: C-tax, L-tax and K-tax denote consumption tax, labor income tax and capital income tax, respectively. 21

22 Table 3. Required consumption tax rate to obtain balanced budget Open economy (W/ borrowing1) No habit persistence (η=0) Open economy (borrowing2) Open economy (borrowing3) Closed economy (no borrowing) C-tax 18.35% 18.39% 18.43% 18.55% Weak habit persistence (η=0.2) C-tax 18.16% 18.20% 18.24% 18.33% Strong habit persistence (η=0.5) C-tax 18.04% 18.06% 18.09% 18.17% Note: 1. The initial steady state consumption tax rate is set at 8.16%. 2. Borrowing 1,2 and 3 indicate unrestricted borrowing, restricted borrowing and very restricted borrowing, respectively. 3. Restricted borrowing (very restricted borrowing) implies that the bond holding over GDP ratio is only at 75% (50%) of that from the unrestricted borrowing case (around 7.6%). 4. Only the consumption tax rate changes, holding the other two tax rates constant, to ensure the intertemporal government budget constraint hold over 5 years. 22

23 Table 4. Welfare implications of a capital income tax cut in both sectors : Lump-sum tax financing No habit persistence (η=0) Overall welfare gains Transitional gains Long-run gains Open economy (borrowing1) Open economy (borrowing2) Open economy (borrowing3) Closed economy (no borrowing) Weak habit persistence (η=0.2) Overall welfare gains 0.17 Transitional gains Long-run gains Strong habit persistence (η=0.5) Overall welfare gains 0.01 Transitional gains Long-run gains Note: 1. Borrowing 1, 2 and 3 indicate unrestricted borrowing, restricted borrowing and very restricted borrowing, respectively. 2. Restricted borrowing (very restricted borrowing) implies that the bond holding over GDP ratio is only at 75% (50%) of that from the unrestricted borrowing case (around 7.6%). 23

24 Table 5. Welfare implications of a capital income tax cut in both sectors : Consumption tax financing No habit persistence (η=0) Overall welfare gains Transitional gains Long-run gains Open economy (borrowing1) Open economy (borrowing2) Open economy (borrowing3) Closed economy (no borrowing) Weak habit persistence (η=0.2) Overall welfare gains 0.16 Transitional gains Long-run gains Strong habit persistence (η=0.5) Overall welfare gains 0.08 Transitional gains Long-run gains Note: 1. Borrowing 1, 2 and 3 indicate unrestricted borrowing, restricted borrowing and very restricted borrowing, respectively. 2. Restricted borrowing (very restricted borrowing) implies that the bond holding over GDP ratio is only at 75% (50%) of that from the unrestricted borrowing case (around 7.6%). 24

25 Table 6. Welfare implications of a capital income tax cut in one sector only : Lump-sum tax financing No habit persistence (η=0) Overall welfare gains Transitional gains Long-run gains Open economy (borrowing1) 0.25 (-0.02) 0.14 (-0.18) 0.11 (0.16) Open economy (borrowing2) 0.19 (-0.05) (-0.37) 0.45 (0.32) Open economy (borrowing3) 0.17 (-0.05) (-0.37) 0.45 (0.32) Closed economy (no borrowing) 0.12 (-0.05) (-0.38) 0.45 (0.32) Weak habit persistence (η=0.2) Overall welfare gains 0.26 (-0.08) Transitional gains 0.17 (-0.22) Long-run gains 0.09 (0.14) 0.14 (-0.14) (-0.46) 0.45 (0.32) 0.11 (-0.14) (-0.46) 0.45 (0.32) 0.04 (-0.15) (-0.47) 0.45 (0.32) Strong habit persistence (η=0.5) Overall welfare gains 0.28 (-0.27) Transitional gains 0.26 (-0.37) Long-run gains 0.02 (0.10) (-0.42) (-0.73) 0.47 (0.31) (-0.43) (-0.74) 0.47 (0.31) (-0.44) (-0.76) 0.47 (0.31) Note: 1. Borrowing 1, 2 and 3 indicate unrestricted borrowing, restricted borrowing and very restricted borrowing, respectively. 2. Restricted borrowing (very restricted borrowing) implies that the bond holding over GDP ratio is only at 75% (50%) of that from the unrestricted borrowing case (around 7.6%). 3. Numbers without the parenthesis are the case of a tax cut in the tradable sector, whereas those in the parenthesis are the case of a tax cut in the non-tradable sector. 25

26 Table 7. Effects on finical balance of capital income tax cut Capital income tax cut in both sectors: lump-sum tax financing Tax base (%) Open Economy Closed Economy C-base L-base K-base K x -base K n -base Tax revenue (%) Dynamic Static Dynamic Static C-rev L-rev K-rev K x -rev K n -rev Total tax revenue Capital income tax cut in the tradable sector (nontradable sector): lump-sum tax financing Tax base (%) Open Economy Closed Economy C-base 1.16 (-0.43) 1.38 (-0.32) L-base 1.86 (0.11) 1.66 (0.01) K-base 1.08 (-0.95) 0.78 (-1.01) K x -base 0.49 (0.49) (0.01) K n -base 1.53 (-2.05) 1.76 (-1.94) Tax revenue (%) Dynamic Static Dynamic Static C-rev 1.16 (-0.43) 0.00 (0.00) 1.38 (-0.32) 0.00 (0.00) L-rev 1.86 (0.11) 0.00 (0.00) 1.66 (0.01) 0.00 (0.00) K-rev (-4.37) (-3.50) (-4.53) (-3.50) K x -rev (0.49) (0.00) (0.01) (0.00) K n -rev 1.53 (-8.09) 0.00 (-6.17) 1.76 (-7.99) 0.00 (-6.17) Total tax revenue 0.84 (-1.15) (-0.91) 0.69 (-1.23) (-0.91) Capital income tax cut in both sectors: consumption tax financing Tax base (%) Open Economy Closed Economy C-base L-base K-base K x -base K n -base Tax revenue (%) Dynamic Static Dynamic Static C-rev L-rev K-rev K x -rev K n -rev Total tax revenue Note: 1. All the results are obtained from a 2.5% permanent tax cut in capital income. 2. C-rev, L-rev, K-rev, K x -rev and K n -rev indicate fiscal revenue from consumption tax, labor income tax, capital income tax in both sectors, capital income tax in the tradable sector and capital income tax in the non-tradable sector, respectively. 3. Dynamic and Static denote dynamic scoring and static scoring. 4. Open (closed) economies indicates unrestricted (no) borrowing and lending with no habit formation. 26