transfers in orer to keep income of the hospital sector unchange, then a larger welfare gain woul be obtaine, even if the government implements a bala

Transcription

1 The Impact of Marginal Tax Reforms on the Supply of Health Relate Services in Japan * Ryuta Ray Kato 1. Introuction This paper presents a computable general equilibrium (CGE) framework to numerically examine the impact of so calle marginal tax reforms of health relate service sectors in Japan. This paper uses the latest Input-Output table of Japan of year 2005 with 108 ifferent prouction sectors, an it explores the effect of marginal changes in tax an subsiy policies from the current situation on economic efficiency, particularly by targeting three important sectors in health relate services; hospital service, social welfare service, an long-term care for the elerly service sectors. (1) The main purpose of this paper is to evaluate the effect on the supply sie of these three service sectors, so that the tax an subsiy of these sectors are particularly consiere. By using the actual input-output table, the paper has successfully realize the real Japanese economy within the moel, an it tries to present welfareenhancing reforms within these three sectors. A istinctive feature of this paper is to focus on the supply sie of these sectors, all of which will play a more important role in aging Japan in the near future. The literature on an aging population of Japan mainly iscusses the effect on public schemes such as the public pension, an public health insurance schemes, since its main concern is with financial burens of population aging on public schemes. It is forecaste not only that population aging will generate more burens on future an working generations through the current schemes, but also that the total number of a population will rastically shrink in the future Japan. A future ecrease in the total population woul likely result in ecreasing future GDP, an stable economic growth of the Japanese economy nees a merging sector to stimulate the economy in an aging Japan. An aging population will inuce more eman for services provie by these three sectors, an these sectors will be more important to stimulate the Japanese economy. Both private an public enterprises in these three sectors have alreay been taxe an subsiize in the current scheme, an the government can thus navigate these sectors with several tax an subsiy policies in orer to achieve higher economic growth. This paper numerically examines the effect of several marginal epartures from the current tax an subsiy policy within a CGE moel, an it explicitly consiers the buget constraint of the government. Any policy change shoul be followe by a seconary policy in orer to fulfill the buget constraint, an this paper tries to present realistic policy scenarios to compensate a sector which will suffer from the policy change. Simulation results are as follows. First of all, an expansion of the subsiy to the hospital sector creates the largest welfare gain when the government oes not take into account its financing explicitly. While such an expansion policy improves economic efficiency, it also inuces a certain amount of government eficits. However, the effect of such a policy on economic efficiency is more than ten times as much as the cost. For instance, the amount of newly generate government eficits is 5.3 billion Japanese yen when the net subsiy rate of the hospital sector increases by 50% from the current level, while the improvement in economic efficiency by the policy is measure to be 72.3 billion Japanese yen. Seconly, however, such an expansion policy oes necessarily not eventuate in the largest gain if the government consiers its balance buget. The reuction of the subsiy to the hospital sector results in the largest welfare gain to the whole economy if the government uses the government surplus inuce by the reuction of the subsiy in orer to ecrease the tax impose on the social welfare sector. When the subsiy to the hospital sector is reuce by 50% from the current level, then the expecte welfare gain to the whole economy woul be approximately 3.8 billion Japanese yen, if the government surplus create by the 50% reuction of the subsiy to the hospital sector is use to reuce the tax on the social welfare sector. In fact, the 50% reuction of the subsiy to the hospital sector eventuates in the social welfare sector being subsiize. Finally, if the hospital sector is compensate by lump-sum transfers when its subsiy is reuce, then a welfare gain coul become larger. If the government uses the government surplus not only for reucing the tax on the social welfare sector but also for proviing the hospital sector with lump-sum 1

2 transfers in orer to keep income of the hospital sector unchange, then a larger welfare gain woul be obtaine, even if the government implements a balance buget policy. When the government reuces the subsiy to the hospital sector by 50% from the current level, the expecte welfare gain to the whole economy is billion Japanese yen. Such a policy keeps the total income of the hospital sector unchange by lump-sum transfers, an also increases the total income of the social welfare sector by reucing its tax. This implies that a welfare enhancing tax reform within the health relate sectors is plausible as long as the subsiy to the hospital sector can be reuce. Such a reform oes not create any new government eficit either. The paper is organize as follows. The next section explains the ata an numerical moel, an Section 3 simulates several scenarios with results an evaluations. Section 4 conclues the paper. 2. Data an Moel This paper employs the conventional static computable general equilibrium (CGE) moel with the actual an the latest input-output table of Japan of year Note that all parameter values in the moel are calculate by using the actual ata, so that the calculate values of enogenous variables obtaine within the moel also become quite realistic. 2.1 Data The latest input-output table of Japan of year 2005 with 108 ifferent intermeiate sectors has been use in orer to construct the social accounting matrix (SAM). The SNA ata has also been use to obtain the amount of aggregate private savings. The last sector, namely the 108th sector, inclues all unclassifie items. Since the value of its factor payments of some intermeiate sectors becomes negative (2), this paper has integrate the 108th sector with the 106th sector which inclues all other services. The integration makes the actual inputoutput table ata consistent to the moel, an it is assume in this paper that there are ifferent prouction sectors, all of which are allowe to have intermeiate prouction processes. Base on this simplification, the social accounting matrix (SAM) has been mae. Then this paper particularly pays more attention to the following three sectors in health relate services; Meical Service an Health (), Social Security (), an Nursing Care (). (3) The main sector in Meical Service an Health () is the hospital sector. Social Security () inclues economic activities of nurseries, nursing homes, social welfare centers, an aministrative work of the public pension as well as public health insurance schemes. Nursing Care () shows economic activities of the inustry of the long-term care for the elerly. Figure 1 shows economic values of omestic final consumption goos of these three sectors in the latest input-output table of year Meical Service an Health () is much larger than other sectors, an its value is 37 thousans billion Japanese yen, while the economic values of other two sectors are between 6.4 thousans billion an 6.6 thousans billion Japanese yen, which are less than 18% of the value of Meical Service an Health () sector. 2.2 Moel The computable general equilibrium moel of this paper employs the conventional static moel. (4) The Japanese economy is assume to consist of ifferent sectors, househols, the government, an the investment firm sector. All inustries are allowe to have intermeiate prouction processes, an they are assume to maximize their profit. Househols are assume to maximize their utility over ifferent consumption goos. The government is assume to etermine its tax revenue, the amount of subsiies, an its consumption in orer to satisfy its buget constraint. The economy is assume to be fully competitive, so that all prices are etermine in the relevant markets in orer to equate the amount of eman to the amount of supply at its fully competitive price level in equilibrium. Note that the moel is static an thus the short-run effect is only investigate. Thus, it is assume for simplicity that factor inputs are not mobile among ifferent sectors in the short-run. <Househols> Househols are assume to be homogenous, an their utility is given by: i =1 U (X 1, X 2,, X ) = Π X i α i, (1) i =1 where X 1 enotes consumption of goo i. i =1α i = 1 is assume. i enotes each sector. The parameter value of each α i is etermine by using the actual social accounting matrix, which is given in Table 6. Househols are assume to maximize (1) with respect to their consumption goos subject to their buget constraint such that: p i X i = I (1 τ I ) S I, 2

3 where p i an I enote the price of goo i an income, respectively. τ I is the proportional income tax rate, an it is calculate by using the actual social accounting matrix. S I enotes the amount of savings, an househols are assume to save the constant amount relative to their isposal income. The amount of savings is assume to be given by S I = s I (1 τ I ) I, where the constant ratio, s I, is given exogenously. (5) The value of s I has been calculate by using the actual SAM. Then income is given by i =1 I = r i K i + w i L i i =1 where r an w enote the rental cost an the wage rate, respectively. K an L are enowments of capital an labour, respectively. The factor payments change as r or w changes. Note that the amounts of r i K i an w i L i are both obtaine from the actual social accounting matrix. The first orer conitions yiel the eman functions such that: X a i I (1 τ I ) (1 s I ) i = X i ( p i, Y; a i ) =, i p i = 1, 2,,. (2) Note that a i can be calculate by using (2) an the actual social accounting matrix so that: a p i X i i = I (1 τ I ) (1 s I ) p i X = i, i (1 s I ) (1 τ I ) ( j =1r j K j + j =1 w j L j ) = 1, 2,,, where both the values of the enominator an the numerator can be obtaine from the actual social accounting matrix. The estimate values of a i are given in Table 6. <Private Firms> Following the conventional assumption, the multiple ecisions by each firm are escribe by the tree structure, where each firm is assume to make a ecision over several ifferent items. In the tree structure, the optimal behavior of each firm which makes a ecision over ifferent items is escribe as if the firm always makes a ecision over two ifferent items at ifferent steps. Each firm makes a ecision over ifferent items; the amount of exports of its own prouct, the amount of importe goos an intermeiate goos use for its prouction, an the amount of labor an capital. This assumption simplifies a complicate ecision over several items by each rm. Each step is also shown in Figure 2. At step 1, a private firm, i, is assume to use labor an capital to prouce its composite goos, Y i. Then, the firm is assume to prouce its omestic goos, Z i, by using its own Y i an Z i, j at the secon step. X i, j enotes the final consumption goos prouce by firm j use by firm i for its prouction. Thus, X i, j is the amount of the final consumption goos prouce by firm j for the intermeiate prouction process of firm i. At the thir step, the firm is assume to ecompose its omestic goos, Z i, into exporte goos, E i, an final omestic goos, D i. This step is concerne about its optimal ecision over the amount of its prouct to be exporte. At the final step (the fourth step), the firm is assume to prouce its final consumption goos, Q i, by using its final omestic goos, D i, an importe goos, M i. This step correspons to its optimal ecision over how much it uses importe goos, M i, an its own goos, D i, to prouce its final consumption goos, Q i, which are consume by omestic househols. The assumption of this tree structure in terms of ifferent ecisions can incorporate firm s complicate ecisions over the amount of exports of its own prouct, the amount of importe goos an intermeiate goos which the firm uses in its prouction process, an the amount of factor inputs into the moel in a tractable way. Note that all market clearing conitions are use to etermine all prices enogenously in their corresponing markets, an also that at each step the private firm is assume to etermine the amount of relevant variables in orer to maximize its profit. By the assumption of the above tree structure, all ecision making processes can be simplifie, an the optimal behavior about all ifferent ecisions can be incorporate as follows: Step 1: The prouction of composite goos Each firm is assume to prouce its composite goos by using capital an labor. Each firm is assume to maximize its profit given by: π i = p i Y Y i (K i, L i ) r i K i w i L i, (3) where Y i an p i Y enote the composite goos prouce by firm i an its price, respectively. K i an L i enote capital an labor use by firm i in orer to prouce its composite goos, respectively. The prouction technology is given by: Y i (K i, L i ) = K i βk, i L i βl, i, i = 1, 2,,, (4) 3

4 where β K, i + β L, i is assume for all i = 1, 2,,. Each firm is assume to maximize (3) with respect to labor an capital subject to (4), an the first orer conitions yiel the eman functions such that: r i K β i K, i =, Y p i Y i β K, i r i K i = K i ( p iy, r i, w i ; β K, i, β L, i ) = p i Y Y i, (5a) L i = L i ( p iy, r i, w i ; β K, i, β L, i ) = = 1, 2,,. (5b) β L, i w i β w i L i L, i =, i = 1, 2,,, Y p i Y i where r i K i, an w i L i can be obtaine from the actual social accounting matrix. The estimate values of β K, i an β L, i are given in Table 6. Z Y X Maxπ i = p i Z i p i Y i p j X i, j, Yi Xi, j j X i, j Y st Z i i = min,, i = 1, 2,,, ax i, j ay i X where X i, j an p j enote intermeiate goo Z use by firm j an i its price, respectively. p i is the price of Z i. ax i, j enotes the amount of intermeiate goo j use for proucing one unit of a omestic goo of firm i, an ay i enotes the amount of its own composite goo for proucing one unit of its omestic goo. The estimate values of ay i are given in Table 6. (6) Note that the prouction function at this step is assume to be the Leontief type. Using ax i, j an ay i, an assuming that the market is fully competitive, the zero-profit conition can be written by: j p i Y Y i, i Note that β K, i an β L, i can be calculate by using (5a), (5b), an the actual social accounting matrix so that: Step 2: The prouction of omestic goos Each firm is assume to prouce omestic goos, Z i, by using intermeiate goos an its own composite goos, which prouction has been escribe at step 1. The optimal behavior of each firm in terms of the prouction of omestic goos can be escribe such that: p i Z = p i Y ay i + p j X ax i, j, i = 1, 2,,. Step 3: Decomposition of Domestic Goos into Exporte Goos an Final Domestic Goos The optimal ecision mae by firm i in terms of the amount of exports of its own goos is escribe as the ecomposition of Z i (i = 1, 2,, ) into exporte goos, E i, an final omestic goos, D i. Each firm is assume to maximize its profit such that: π i = p i e E i + p i D i (1+ τ i p τ i s ) p i Z Z i, (6) where p i e an p i enote the price when the omestic goos are sol abroa, an the price when the omestic goos are sol omestically, respectively. Note that p i e is measure in the omestic currency. τ i p an τ i s are the tax rates of a prouction tax impose on the prouction of Z i an the subsiy rate, respectively. The values of τ i p an τ i s are calculate by using the actual social accounting matrix, an the calculate values are given in Table 2-1 an 2-2. The ecomposition is assume to follow the Cobb-Douglas technology such that: e i ii Z i = E i κ D i κ, i = 1, 2,,, (7) where κ i + κ i e = 1 (i = 1, 2,, ) is assume. Each firm is assume to maximize (6) with respect to E i an D i subject to (7), an the first orer conitions yiel E i = E i ( p ie, p i, p iz ; τ i p, τ i s, κ i, κ ie ) = κ i e (1+ τ i p τ i s ) p i Z Z i p i e κ i (1+ τ i p τ i s ) p i Z Z i p i, (8a) D i = D i ( p ie, p i, p iz ; τ i p, τ i s, κ i, κ ie ) = e e p i E κ i i = (1+ τ p s, Z i τ i ) p i Z i, i = 1, 2,,, (8b) Note that κ i e an κ i can be calculate by using (8a), (8b), an the actual social accounting matrix so that: p i D κ i i = (1+ τ p s Z, i = 1, 2,,, i τ i ) p i Z i where p i e E i, p i D i, τ i s p i Z Z i, an τ i p p i Z Z i can be obtaine from the actual social accounting matrix. The estimate values of κ i e an κ i are given in Table 6. 4

5 Step 4: The Prouction of the final goos Denote the final consumption goos by Q i (i = 1, 2,, ). The final consumption goos are assume to be prouce by using the final omestic goos, D i, an the importe goos, M i. This step correspons to the optimal ecision making behavior of each firm in terms of the amount of importe goos which are use in its prouction process. The prouction technology at this final step is given by the following Cobb-Douglas function: Q i = M i γ m γ i D i i, i = 1, 2,,, (9) where γ i m + γ i = 1 (i = 1, 2,, ) is assume. Each firm is assume to maximize its profit with respect to M i an D i subject to (9). Its profit is given by: π i = p i Q Q i (1+τ i m ) p i m M i p i D i, i = 1, 2,,, where p i Q an τ i m enote the price of its final consumption goos, Q i, an the import tariff rate, respectively. The import tariff rate is calculate by using the actual social accounting matrix, an it is given in Table 2-4. Then, the first orer conitions yiel M i = M i ( p im, p i, p iq ; τ im, γ im, γ i ) m Q = γ i p i Q i m, (10 a) m (1+ τ i ) p i D i = D i ( p im, p i, p iq ; τ im, γ im, γ i ) Q γ i p i Q = i, i = 1, 2,,, (10 b) p i Note that γ i m an γ i can be calculate by using (10a), (10b), an the actual social accounting matrix so that: m m m (1+ τ i ) p i M γ i i =, Q p i Q i p i D γ i i =, i = 1, 2,,, Q p i Q i Q m where p im M i, p i D i, p i Q i an τ i p im M i can be obtaine from the actual social accounting matrix. m The estimate values of γ i an γ i are given in Table 6. <The Government> The government is assume to impose several taxes to satisfy its buget constraint. Its buget constraint is given by: p i Q X i g + S g + Sub = T I + T p + T m, i =1 where the left han sie is the total government expeniture, an the right han sie is the total government revenue. X i g an S g enote government consumption of final consumption goo i; an government savings, respectively. Sub enotes the total amount of subsiies such that: Sub = τ i s ( p i z Z i ). i =1 The total tax revenue is given by: T I = τ I I = τ I r i K i + w i L i =1 i =1 i, T p p = τ i ( z p i Z i ), i =1 T m = τ i m ( p i m M i ), i =1 where T I, T p an T m enote the total income tax revenue, the total prouction tax revenue, an the total import tariff revenue, respectively. The government is assume to save the constant amount relative to the total amount of tax revenue, an the government savings are assume to be given by S g = s g ( T I + T p + T m ), where the constant ratio, S g, is given exogenously, an.its value has been calculate by using the actual SAM. <Equilibrium Conitions> There are two factor inputs, labour an capital. Since the moel is static an thus the short-run effect is explore, it is assume that each factor cannot move among ifferent sectors (inustries) in the short-run. This implies the equilibrium conitions of factor markets such that K i = K i, (11 a) L i = L i, i = 1, 2,,, (11 b) where the total amount of enowments is given by: K = K i, i =1 L = L i, i =1 Note that r i an w i (i = 1, 2,, ) are etermine in orer to satisfy (11a) an (11b), respectively. In terms of the market clearing conition of goo i (i = 1, 2,, ); a private investment sector is introuce in orer to close the economy in this 5

6 paper. (7) Denoting the amount of goo i consume by the private investment sector by X is, the market clearing conition of goo i is given by: i =1 i =1 i =1 Q i = X i + X i g + X i s + X ij,i = 1, 2,,, (12) j where the left han sie is the total supply, an the right han sie is the total eman for goo i. p i Q (i = 1, 2,, ) is etermine in orer to satisfy (12). Note that the buget constraint of the private investment sector is given by: p i Q X i s = S g + S I + S f, where the left han sie is the total amount of its consumption, an the right han sie is the total amount of its income. S f enotes the total amount of savings by the foreign sector, or the eficits in the current account, an it is given by subtracting exports from imports. (8) Since both the amount of exports an the amount of imports can be obtaine from the actual social accounting matrix, S f can be calculate from the actual social accounting matrix, an thus it is exogenously given in the moel. Furthermore, the foreign trae balance is given by p i w, e E i + S f = p i w, m M i, where p i w, e an p i w, m enote the worl price of export goos, an import goos of goo i, respectively, an both of them are assume to be given exogenously. Since p i e an p i m are both measure in the omestic currency, they are also expresse such that: p i e = ε p i w, e, p i m = ε p i w, m, i = 1, 2,,, where ε enotes the exchange rate. Note that the exogeneity assumption on the worl prices implies that the exchange rate is enogenously etermine within the moel. 3. Simulation Analysis 3.1 Benchmark an Calibration The benchmark case shoul reflect the real Japanese economy in orer to make the subsequent simulation scenarios realistic. Thus, the benchmark moel shoul carefully be calibrate until the calculate values of all enogenous variables within the moel become close to the actual values. Table 1-1 to 1-4 show the calculate moel values as well as the corresponing actual values in year As shown in these tables, the benchmark case has successfully been able to reprouce the real economy within the moel. Note that the tax rates an the subsiy rates shown in Table 2-1 to 2-4 have been calculate by using the actual amount of tax an subsiy, so that they can be interprete as the average proportional rates. Table 2-3 particularly shows the net rate, which is efine as the ifference between the prouction tax rate an the subsiy rate. The negative value of the net rate implies that the sector is subsiize by a certain amount. (9) As Table 2-3 shows, only Meical Service an Health (: hospital sector) is subsiize (net subsiy rate: %) among all relevant three sectors. Since the effect of changes in the net rate is only simulate in the subsequent sections, the net rates of these three sectors are shown again in Table Note also that welfare gains in the next section are all measure by equivalent variation (EV), so that the effect of policy changes on economic efficiency are measure financially. 3.2 Simulations Scenario I without balance buget Any policy change shoul be followe by a seconary policy if the buget constraint of the government is fulfille even after the policy change. In this paper, the total government expeniture is assume to be unchange even after a policy change. This implies that the total revenue shoul be unchange in orer to fulfill the buget constraint, so that any policy change shoul be followe by a seconary policy in orer to satisfy the buget constraint. However, a seconary policy conucte in orer to fulfill the buget constraint obviously generates another effect on an economy, so that it is very ifficult to separate the obtaine result into the effects of the first an secon policies, respectively. Then, Table 3-1 shows the pure effect of a policy change without a seconary policy base on the assumption that the gap between revenue an expeniture cause by any policy change is finance by government bons implicitly. While the buget is not balance after a policy change, it can show how much the Japanese government nees to conuct marginal tax/subsiy reforms. Note that Table 3-1 shows to the extent how much economic gains woul be obtaine by changing the net rate of each sector from the current level uner the assumption that the temporary buget is not balance. The negative value of government eficits in the table implies that government surplus will be generate by a policy change. Since the economic size of Meical Service an Health (: hospital sector) is much larger than other 6

7 two sectors, an increase in the subsiy to Meical Service an Health (: hospital sector) results in the largest welfare gain. For instance, if the government increases the net subsiy rate of this sector by 50% from the current level, then the expecte gain in economic efficiency is measure to be 72.3 billion Japanese yen with newly generate government eficits of approximately 5.6 billion Japanese yen. However, note that the overall effect of such a policy is more than ten times as much as the cost, since the amount of newly generate government eficits (5.3 billion Japanese yen) is less than one tenth of a welfare gain (72.3billion Japanese yen) Scenario II with balance buget In this scenario, the buget constraint is explicitly fulfille with a seconary tax/subsiy policy. While it is possible to consier many seconary policies to fulfill the buget constraint, it is assume that the net rates of three sectors are only consiere. Note that both the prouction tax an the subsiy are istortionary. Since the net rate is moifie from the current level, the environment consiere in this paper is the secon-best situation, implying that the overall effect on economic efficiency might be positive or negative as pointe out by Lipsey an Lancaster (1956). The sector, which net rate is moifie exogenously, is calle the initial sector, an the sector, which net rate is ajuste enogenously in orer to fulfill the buget constraint, is calle the seconary sector. Table 3-2 shows the overall effect of such policies. A striking result is that when the buget is balance by a seconary policy the result is quite ifferent. If the balance buget is not consiere explicitly, the most effective welfare enhancing policy is to more subsiize Meical Service an Health (: hospital sector).however, if the gap between revenue an expeniture cause by the initial policy change is finance by a istortionary tax/subsiy policy within the health relate sectors, then the reuction of the subsiy to Meical Service an Health (: hospital sector) is more preferable oppositely. While a policy to expan the subsiy to Meical Service an Health (: hospital sector) still eventuates in a welfare gain irrespective of a seconary policy, a welfare gain generate by a policy to reuce the subsiy followe by a seconary policy to ajust the net rate of Social Security () is the largest. When the net subsiy rate of Meical Service an Health (i = 94: hospital sector) is reuce by 50% from the current level, the expecte welfare gain is 3.78 billion Japanese yen if the policy is followe by the enogenous ajustment of the net rate of Social Security () sector. As long as the effect on economic efficiency of the whole economy with the balance buget is concerne, a reform with the reuction of the subsiy to Meical Service an Health (: hospital sector) sector followe by an enogenous tax cut in Social Security () sector is most effective. However, such a policy results in a ecrease (an increase) in the total income of Meical Service an Health (: hospital sector) sector. Table 4-1 shows that while the total income of Social Security () sector is expecte to increase by % from the current level, that of Meical Service an Health (: hospital sector) sector is expecte to ecrease by %. Table 4-2 also shows that such a policy eventuates 17 in Social Security () sector being subsiize. This is because the economic size of Meical Service an Health (: hospital sector) sector is much larger than that of Social Security (i = 95) sector, so that a 50% reuction of the subsiy rate of Meical Service an Health (: hospital sector) sector inuces a huge amount of the government surplus, resulting in the government subsiizing Social Security () sector. Such a policy is obviously favorable for Social Security () sector, but it is not for Meical Service an Health (: hospital sector) sector. Since the economic size of Meical Service an Health (: hospital sector) is quite large, it seems politically ifficult to implement such a policy. Then in the next scenario, a compensation policy is investigate Scenario III with balance buget an a compensation policy Scenario II shows that the reuction of the subsiy to Meical Service an Health (: hospital sector) sector followe by a ecreasing tax policy of Social Security () sector results in the largest welfare gain to the whole economy. Thus, Scenario III only investigates the case where the net tax rate of Social Security () sector is enogenously moifie as a seconary policy in orer to full the buget constraint. Furthermore, Scenario III assumes that Meical Service an Health (: hospital sector) sector is compensate by lump-sum transfers, so that the total income of Meical Service an Health (: hospital sector) sector keeps unchange even after an exogenous ecrease in its net subsiy rate. Table 5 shows the striking result. When the net subsiy rate of Meical Service an Health (i = 94: hospital sector) sector is reuce by 50% from the current level, a welfare gain to the whole economy is measure to be billion Japanese 7

8 yen, which is much larger than the case where all government surplus generate by the reuction of the subsiy is use to ecrease the tax on Social Security () sector. If the surplus is only use for the reuction of the tax on Social Security (i = 95) sector, then expecte amount of a welfare gain is only 3.78 billion Japanese yen as shown in Table 3-2. While the net tax rate for Social Security () sector is higher in Scenario III compare to Scenario II, it still obtains an increase in its income, since its net rate can be reuce by such a policy. This implies that it is plausible to enhance welfare (economic efficiency) even in the health relate sectors if the amount of the subsiy to Meical Service an Health (: hospital sector) sector can be reuce. Note that Scenario III uses non-istortionary lump-sum transfers in orer to compensate Meical Service an Health (: hospital sector) sector. In Scenario III, the istortionary subsiy to Meical Service an Health (: hospital sector) sector an the istortionary tax on Social Security () sector are both reuce, an the government surplus is reistribute to Meical Service an Health (i = 94: hospital sector) sector by lump-sum transfers. Note also that the amount of lump-sum transfers is more than three times as much as a welfare gain to the whole economy. This implies that the government has to reistribute lots of resources through transfers to improve economic efficiency if the government tries to reform health relate prouction sectors. 4. Concluing Remarks This paper has presente a computable general equilibrium (CGE) framework to numerically examine the effect of marginal tax reforms on the supply sie of health relate sectors in Japan. This paper has use the latest Input-Output table of Japan of year 2005 with ifferent prouction sectors. Several simulations have been conucte in comparison with a very realistic benchmark moel, an the obtaine results are as follows. First of all, an expansion of the subsiy to the hospital sector creates the largest welfare gain when the government oes not take into account its financing explicitly. While such an expansion policy improves economic efficiency, it also inuces a certain amount of government eficits. However, the effect of such a policy on economic efficiency is more than ten times as much as the cost. For instance, the amount of newly generate government eficits is 5.3 billion Japanese yen when the subsiy to the hospital sector increases by 50% from the current level, while the improvement in economic efficiency by the policy is measure to be 72.3 billion Japanese yen. Seconly, however, such an expansion policy oes necessarily not eventuate in the largest gain if the government consiers its balance buget. The reuction of the subsiy to the hospital sector results in the largest welfare gain to the whole economy if the government uses the government surplus inuce by the reuction of the subsiy in orer to ecrease the tax impose on the social welfare sector. When the subsiy to the hospital sector is reuce by 50% from the current level, then the expecte welfare gain to the whole economy woul be approximately 3.8 billion Japanese yen, if the government surplus create by the 50% reuction of the subsiy to the hospital sector is use to reuce the tax on the social welfare sector. In fact, the 50% reuction of the subsiy to the hospital sector eventuates in the social welfare sector being subsiize. Finally, if the hospital sector is compensate by lump-sum transfers when its subsiy is reuce, then a welfare gain coul become larger. If the government uses the government surplus not only for reucing the net tax rate of the social welfare sector but also for proviing the hospital sector with lump-sum transfers in orer to keep income of the hospital sector unchange, then a larger welfare gain woul be obtaine, even if the government implements a balance buget policy. When the government reuces the subsiy to the hospital sector by 50% from the current level, the expecte welfare gain to the whole economy is billion Japanese yen. Such a policy keeps the total income of the hospital sector unchange by lump-sum transfers, an also increases the total income of the social welfare sector by reucing its tax. This implies that a welfare enhancing tax reform within the health relate sectors is plausible as long as the subsiy to the hospital sector can be reuce. Such a reform oes not create any new government eficit either. While this paper has use the Japanese inputoutput table, it is applicable to all other countries in orer to investigate the effect of several health policies. By explicitly consiering the buget constraint within a computable general equilibrium framework, this paper has thrown light on the importance of explicit consieration of the government buget constraint when simulations on tax an subsiy policies are conucte. Ryuta Ray Kato, Ph.D (Grauate School of International Relations) 8

9 Notes *) Keywors: Computable General Equilibrium (CGE) Moel, Marginal Tax Reform, Health Sectors, Taxation, Subsiy, Simulation; JEL Classification: C68, H51, an H53. 1) Kato (2011) also iscusses the effect of reforms on the supply sie of the pharmaceutical inustry. 2) Labor income an capital income are factor payments. 3) The numbers in the brackets are numbers allocate to the sectors in the actual inputoutput table of year 2005 with 108 ifferent prouction sectors. 4) In terms of the conventional static moel, see Ballar, Fullerton, Shoven, an Whalley (1985), Shoven an Whalley (1992), an Scarf an Shoven (2008). In particular, the moel use in this paper is similar to Hosoe, Ogawa, an Hashimoto (2004). Regaring the ynamic moel, it is conventional to employ an overlapping generations moel In terms of computable overlapping generations moel within a general equilibrium framework, see Auerbach an Kotlikoff (1987). Kato (1998), Kato (2002b), Kato (2002a), Ihori, Kato, Kawae, an Bessho (2006), an Ihori, Kato, Kawae, an Bessho (2011) also apply the yanamic moel to several policies in Japan. 5) The assumption that the ratio is exogenously given is mae only for the moel to be consistent to the actual social accounting matrix, an this assumption is very common in the literature. 6) The estimate values of ax ij are not presente in Table 2, since the number of the estimate values reach 11,449. The estimate values are given upon request. 7) This is also the conventional assumption in the literature. 8) The FDI is assume to be negligible in this paper. 9) A tariff is ifferently treate, so that the net rate is efine above. References Auerbach, A. J., an L. J. Kotlikoff (1987): Dynamic Fiscal Policy. Cambrige University Press. Ballar, C. L., D. Fullerton, J. B. Shoven, an J. Whalley (1985): A General Equilibrium Moel for Tax Policy Evaluation. Chicago University Press. Hosoe, N., K. Ogawa, an H. Hashimoto (2004): Textbook of Computable General Equilibrium Moeling. University of Tokyo Press. Ihori, T., R. R. Kato, M. Kawae, an S. Bessho (2006): Public Debt an Economic Growth in an Aging Japan, in Tackling Japan s Fiscal Challenges, e. by K. Kaizuka, an A. O. Krueger. Palgrave. (2011): Health Insurance Reform an Economic Growth: Simulation Analysis in Japan, Japan an the Worl Economy, Forthcoming. Kato, R. R. (1998): Transition to an aging Japan: Public pension, savings, an capital taxation, Journal of the Japanese an International Economies, 12, (2002a): Government eficit, public investment, an public capital in the transition to an aging Japan, Journal of the Japanese an International Economies, 16, (2002b): Government eficits in an aging Japan, in Government Deficit an Fiscal Reform in Japan, e. by T. Ihori, an M. Sato. Kluwer Acaemic Publishers. (2011): Who Has to Pay More, Health Service Sectors, the Pharmaceutical Inustry, or Future Generations?: A Computable General Equilibrium Approach, Economics an Management Series, IUJ Research Institute, EMS Lipsey, R. G., an K. Lancaster (1956): The General Theory of Secon Best, The Review of Economic Stuies, 24(1), Scarf, H. E., an J. B. Shoven (es.) (2008): Applie General Equilibrium Analysis. Cambrige University Press. Shoven, J. B., an J. Whalley (1992): Applying General Equilibrium. Cambrige University Press. 9

10 40,000 35,000 30,000 25,000 20,000 15,000 10,000 5,000 0 Figure 1: Economic Values of the Domestic Final Consumption Goos in the IO Table of Year ,209 Unit: One billion Japanese yen 6,616 6,388 : Meical Service an Health Sector (incl. hospitals) : Social Security Sector (incl. nurseries, nursing homes, social welfare centers, an aministrative work of public schemes) : Nursing Care Sector (incl. long term care for the elerly) 10

11 Table 1-1: Economic Values of the Benchmark Moel Final Consumption Goos, P i Q Q i ; i = 1, 2,, Unit: One million Japanese yen i moel actual i moel actual i moel actual i moel actual i moel actual i moel actual i moel actual i moel actual i moel actual i moel actual i moel actual i moel actual : Meical Service an Health Sector (incl. hospitals) : Social Security Sector (incl. nurseries, nursing homes, social welfare centers, an aministrative work of public schemes) : Nursing Care Sector (incl. long term care for the elerly) 11

12 Table 1-2: Economic Values of the Benchmark Moel (Continue) Capital Income, rk i ; i = 1, 2,, Unit: One million Japanese yen i moel actual i moel actual i moel actual i moel actual i moel actual i moel actual i moel actual i moel actual i moel 2.50E E E actual 2.50E E E i moel actual i moel 1.20E actual 1.20E i moel actual : Meical Service an Health Sector (incl. hospitals) : Social Security Sector (incl. nurseries, nursing homes, social welfare centers, an aministrative work of public schemes) : Nursing Care Sector (incl. long term care for the elerly) 12

13 Table 1-3: Economic Values of the Benchmark Moel (Continue) Labor Income, wl i ; i = 1, 2,, Unit: One million Japanese yen i moel actual i moel actual i moel actual i moel actual i moel actual i moel actual i moel actual i moel E actual E i moel 4.40E E actual 4.40E E i moel actual i moel 1.70E E E actual 1.70E E E i moel E actual E : Meical Service an Health Sector (incl. hospitals) : Social Security Sector (incl. nurseries, nursing homes, social welfare centers, an aministrative work of public schemes) : Nursing Care Sector (incl. long term care for the elerly) 13

14 Table 1-4: Economic Values of the Benchmark Moel (Continue) Unit: One million Japanese yen savings private sector government sector foreign sector moel actual moel actual moel actual tax an subsiy income tax prouction tax import tax subsiy moel actual moel actual moel actual moel actual The above figures inicate the total amount. Table 2-1: Calculate Prouction Tax Rates TAUP (i) = τ ip ; i = 1, 2,, (Subsiy Rate) TAUP ( 1) TAUP ( 2) TAUP ( 3) TAUP ( 4) TAUP ( 5) TAUP ( 6) TAUP ( 7) TAUP ( 8) TAUP ( 9) % % % % % % % % % TAUP ( 10) TAUP ( 11) TAUP ( 12) TAUP ( 13) TAUP ( 14) TAUP ( 15) TAUP ( 16) TAUP ( 17) TAUP ( 18) % % % % % % % % % TAUP ( 19) TAUP ( 20) TAUP ( 21) TAUP ( 22) TAUP ( 23) TAUP ( 24) TAUP ( 25) TAUP ( 26) TAUP ( 27) % % % % 2.8% % % % % TAUP ( 28) TAUP ( 29) TAUP ( 30) TAUP ( 31) TAUP ( 32) TAUP ( 33) TAUP ( 34) TAUP ( 35) TAUP ( 36) % % % % % % % % % TAUP ( 37) TAUP ( 38) TAUP ( 39) TAUP ( 40) TAUP ( 41) TAUP ( 42) TAUP ( 43) TAUP ( 44) TAUP ( 45) % % % % % % % % % TAUP ( 46) TAUP ( 47) TAUP ( 48) TAUP ( 49) TAUP ( 50) TAUP ( 51) TAUP ( 52) TAUP ( 53) TAUP ( 54) % % % % % % % % % TAUP ( 55) TAUP ( 56) TAUP ( 57) TAUP ( 58) TAUP ( 59) TAUP ( 60) TAUP ( 61) TAUP ( 62) TAUP ( 63) % % % % % % % % % TAUP ( 64) TAUP ( 65) TAUP ( 66) TAUP ( 67) TAUP ( 68) TAUP ( 69) TAUP ( 70) TAUP ( 71) TAUP ( 72) % % % % % % % % % TAUP ( 73) TAUP ( 74) TAUP ( 75) TAUP ( 76) TAUP ( 77) TAUP ( 78) TAUP ( 79) TAUP ( 80) TAUP ( 81) % % % % % % % % % TAUP ( 82) TAUP ( 83) TAUP ( 84) TAUP ( 85) TAUP ( 86) TAUP ( 87) TAUP ( 88) TAUP ( 89) TAUP ( 90) % % % % % % % % % TAUP ( 91) TAUP ( 92) TAUP ( 93) TAUP ( 94) TAUP ( 95) TAUP ( 96) TAUP ( 97) TAUP ( 98) TAUP ( 99) % % % % % % % % % TAUP (100) TAUP (101) TAUP (102) TAUP (103) TAUP (104) TAUP (105) TAUP (106) TAUP () % % % % % % % % : Meical Service an Health Sector (incl. hospitals) : Social Security Sector (incl. nurseries, nursing homes, social welfare centers, an aministrative work of public schemes) : Nursing Care Sector (incl. long term care for the elerly) 14

15 Table 2-2: Calculate Subsiy Rates SUBR (i) = τ is ; i = 1, 2,, (Subsiy Rate) SUBR ( 1) SUBR ( 2) SUBR ( 3) SUBR ( 4) SUBR ( 5) SUBR ( 6) SUBR ( 7) SUBR ( 8) SUBR ( 9) % % % % % % % % % SUBR ( 10) SUBR ( 11) SUBR ( 12) SUBR ( 13) SUBR ( 14) SUBR ( 15) SUBR ( 16) SUBR ( 17) SUBR ( 18) % % % % % % % % % SUBR ( 19) SUBR ( 20) SUBR ( 21) SUBR ( 22) SUBR ( 23) SUBR ( 24) SUBR ( 25) SUBR ( 26) SUBR ( 27) % % % % % % % % % SUBR ( 28) SUBR ( 29) SUBR ( 30) SUBR ( 31) SUBR ( 32) SUBR ( 33) SUBR ( 34) SUBR ( 35) SUBR ( 36) % % % % % % % % % SUBR ( 37) SUBR ( 38) SUBR ( 39) SUBR ( 40) SUBR ( 41) SUBR ( 42) SUBR ( 43) SUBR ( 44) SUBR ( 45) % % % % % % % % % SUBR ( 46) SUBR ( 47) SUBR ( 48) SUBR ( 49) SUBR ( 50) SUBR ( 51) SUBR ( 52) SUBR ( 53) SUBR ( 54) % % % % % % % % % SUBR ( 55) SUBR ( 56) SUBR ( 57) SUBR ( 58) SUBR ( 59) SUBR ( 60) SUBR ( 61) SUBR ( 62) SUBR ( 63) % % % % % % % % % SUBR ( 64) SUBR ( 65) SUBR ( 66) SUBR ( 67) SUBR ( 68) SUBR ( 69) SUBR ( 70) SUBR ( 71) SUBR ( 72) % % % % % % % % % SUBR ( 73) SUBR ( 74) SUBR ( 75) SUBR ( 76) SUBR ( 77) SUBR ( 78) SUBR ( 79) SUBR ( 80) SUBR ( 81) % % % % % % % % % SUBR ( 82) SUBR ( 83) SUBR ( 84) SUBR ( 85) SUBR ( 86) SUBR ( 87) SUBR ( 88) SUBR ( 89) SUBR ( 90) % % % % % % % % % SUBR ( 91) SUBR ( 92) SUBR ( 93) SUBR ( 94) SUBR ( 95) SUBR ( 96) SUBR ( 97) SUBR ( 98) SUBR ( 99) % % % % % % % % % SUBR (100) SUBR (101) SUBR (102) SUBR (103) SUBR (104) SUBR (105) SUBR (106) SUBR () % % % % % % % % : Meical Service an Health Sector (incl. hospitals) : Social Security Sector (incl. nurseries, nursing homes, social welfare centers, an aministrative work of public schemes) : Nursing Care Sector (incl. long term care for the elerly) 15