Tax Reform Issues in Korea*
Junghun Kim Korea Institute of Public Finance 79-6 Garak-Dong, Songpa-Gu, Seoul 138-774, Korea Email: [email protected]
-----------------------------------------------This is a paper to be presented at “International Symposium on Tax Policy and Reform in Asia,” Hitotsubashi University, July 1-2, 2005.
Table of Contents I. Introduction ...................................................................................................................... II. Overview of Taxation in Korea ...................................................................................... 1. Tax system................................................................................................................... 2. Tax revenue ................................................................................................................. 3. Rate structures ............................................................................................................. III. Tax Reform Issues 1. Economic growth......................................................................................................... 2. Tax competition and FDI............................................................................................. 3. Redistribution .............................................................................................................. 4. Political Economy........................................................................................................ 5. Tax assignment ............................................................................................................ 6. Taxes on real estate...................................................................................................... IV. Conclusion..................................................................................................................... References ...........................................................................................................................
I. Introduction Taxation and expenditure policies in Korea have been relatively simple issues. The burden of taxation, often a central political issue in developed countries, has not attracted much attention from voters and politicians since political parties in Korea are not clearly divided on the taxation issue. As a matter of fact, all political parties in Korea claim that they are for middle-classes and working-classes. Due to rather unsophisticated attitude toward taxes, lowering tax rates of income tax has been generally regarded as a good tax policy even by low income groups.
The fact that the tax burden is not a politically sensitive issue is reflected in the structure of the tax burden. The level of the tax burden in Korea is not only among the lowest in OECD countries, but also the burden of income tax is among the lowest. On the other hand, the revenue from the direct taxes such as VAT and gasoline tax occupies more than 50 percent of the total tax revenue.
As for the expenditure side, the principle of balanced budget was virtually in effect until 1998, the year when financial crisis hit the Korean economy. Before the financial crisis and restructuring process forced the government to have a sizable amount of government debt, the need for raising tax revenue was not a great concern for the government, partly because of the high growth rates of GDP and tax revenue, and partly because the amounts of social expenditures such as welfare programs and health care were limited. Therefore low tax burden and expenditure within the budget were two main characteristics that defined public finance in Korea until late 1990s.
Taxation also has played a limited role in the process of decentralization. Even if Korea started local autonomy in 1995, few citizens pay attention to the differences between national and local taxes. This is because local governments do not rely on local taxes to raise revenue. It is often argued that local governments in Korea do not have an independent power to tax, but that is not correct since the local tax law passed by the parliament allows local governments to set local tax rates within certain boundaries, usually plus/minus 50 percent. Therefore, the reason why local taxes play a limited role
in local decision making should be found in the incentive structure of local public finance in Korea. But more fundamentally, taxes, both central and local, are not the main tools for voters to represent their interests, since politics in Korea are more influenced by other factors than fiscal policies. 1 The relatively simple nature of the fiscal policy environment in Korea seems to be changing recently, however. After the financial crisis in the late 1990s, the nature of Korean economy has changed in many ways. The number of the unemployed increased sharply during the restructuring process, and in response to that, social safety net has been expanded. The public pension program was also significantly extended in 1999 to cover virtually all citizens. The financing structure of companies has changed also. Before the financial crisis, companies heavily depended on debt financing, but the amount of corporate debt has been significantly reduced since then. The changed financing structure of companies implies that interest expensing has been also reduced, and it has contributed to a record high increase of corporate tax revenue for the past several years. While it is true that the tax base of the corporate tax has expanded recently, its long-term role as a revenue source seems to be shrinking, rather than expanding. This is because the corporate tax rates have been decreased recently, and a further cut of the tax rates is continuously being discussed. The reason why the corporate tax rates are under downward pressure is partly because the corporate tax revenue has been recently increasing very fast. However, another important reason is the possibility of tax competition. The issue of tax competition has not attracted as strong attention as in the European countries or United States. However, the concern over tax competition is a noticeable element in recent tax policy debates. The international comparison of corporate tax rates, which are generally decreasing, is often used as a policy benchmark in the debates over the level of corporate tax rates in Korea. If international comparison provides any indication to the tax policy in Korea, it will be 1
Recently, there have been exceptional cases. The new government significantly raised the assessed value of properties in order to reflect the rapidly increasing real estate prices. In response to this, the local governments in rich jurisdiction, mostly located in the Seoul metropolitan region, decreased the local tax rates of the property tax. This episode is interesting since it involves the citizens’ responsiveness to tax policies, tax and expenditure assignments between the central and local governments. More will be
the weak role of the individual income tax since its rates are among the lowest in OECD countries. However, other considerations are more important than international comparison in the debates over income tax policy in Korea. The first is the issue of horizontal equity between employees and self-employed. Because employees strongly complain about their relatively high income tax burden, the overall level of income tax rates has been kept at a low level to avoid political controversies. Another factor that limits the role of income tax is the contributions to pension and health care program that have increased very rapidly since late 1990s. These are important factors that constrain the role of income tax in Korea. On the other hand, the income gap is widening and the income tax burden, with social security included, is among the lowest in OECD countries. Therefore it seems that there is still a room for bigger role of income tax in Korea. However, because of the unpopular view toward income tax, any attempt to increase income tax burden will be faced with strong resistance from middle and high-income individuals. In contrast to income tax, the taxes on real estates are highly progressive and attract considerable political attention. This is because the real estate prices have been rising fast for the past 30 years, resulting in a very skewed distribution of wealth, especially real estates. Recently, the prices of the real estates, especially Seoul metropolitan areas, have again sharply increased, and heated debates on the role of the taxes on real estates are undergoing. The structures of taxes on real estates are rather unique in Korea, since the recurrent property tax has very progressive tax rates, and majority of real estates tax revenue come from taxes on acquisition and registration of real estates. The paper is organized as follows. an overview of taxation in Korea is in section II. In section 3, the tax reform issues in Korea are discussed with focuses on economic growth, FDI, redistribution, tax assignment, and real estates. Section 4 concludes the paper.
discussed in the later part of this paper.
II. Overview of Taxation in Korea 1. Tax system The Korean tax system consists of fourteen national taxes and sixteen local taxes. The national tax is categorized into “internal taxes,” customs duties, and earmarked taxes. The internal taxes consist of ten taxes. 2 The definition of the internal taxes is important in understanding Korea’s fiscal structure since 19.13% of the internal taxes is the basis of general grants to local governments. Another 19.4% of the internal taxes is the basis of education general grants. Therefore 36.53% of the internal taxes is transferred as general grants to both local governments and the education special account. 3 The earmarked taxes are not the basis of general grants, and consist of Transportation Tax, Education Tax, and Special Tax for Rural Development. 4 This structure creates an incentive for the central government to prefer earmarked taxes: when 1 Won of earmarked tax is collected, 1 Won belongs to the central government's revenue, while only 0.63 Won of internal taxes belongs to the central government's revenue. Among the fourteen national taxes in Korea, thirteen taxes are the common types of taxes that are found in many other countries, except the Comprehensive Real Estate Holding Tax. This tax was created in 2005 and has a distinct feature. Previously, there were two types of recurrent property taxes, both of which being local taxes: Aggregate Land Tax was levied on land, and Property Tax on structures. Unlike other countries, recurrent property taxes in Korea have highly progressive rate structures and it was decided in 2004 that the central government takes the upper bracket of the progressive tax rates. It was also decided that the tax base of the recurrent property tax would be the assessed value of a property, rather than separate assessments of land and structures. As a result, both the central and local governments collect the recurrent property tax: local governments apply the lower brackets of the progressive tax rates (Property Tax), and
2 These are Income Tax, Corporation Tax, Inheritance Tax, Gift Tax, and Comprehensive Real Estate Holding Tax (direct taxes), Value-added Tax, Special Excise Tax, Liquor Tax, Stamp Tax, and Securities Transaction Tax (indirect taxes). 3 Education is not a part of local public services provided by local governments in Korea. Education special account is managed by the Ministry of Education. 4 In the earmarked taxes, Transportation Tax, a tax on fuels, is an indirect tax. Education Tax and Special Tax for Rural Development are a special kind of surtaxes that are piggybacked on many other tax items. Education Tax is a surtax on four national taxes and seven local taxes. Special Tax for Rural Development is a surtax on 2 national taxes, and 3 local taxes.
the central government applies the upper brackets of the progressive tax rates (Comprehensive Real Estate Holding Tax). The local taxes are collected by metropolitan cities, special districts, provinces and cities/towns. 5 The metropolitan cities (Seoul and other six cities) collect fourteen local taxes, 6 and the special districts collect three local taxes. 7 Among the sixteen local taxes, three taxes were recently introduced. As was explained previously, Property Tax was introduced in 2005. In 2000, Local Fuel Tax was introduced as a shared tax with the central government’s Transportation Tax. In 2001, Local Education Tax was introduced. The nature of Local Education Tax is, however, rather ambiguous. Before 2001, the Education Tax was piggybacked on four national taxes, and seven local taxes and then the revenue was transferred to an education special account managed by the Ministry of Education. As of 2001, the surtaxes on local taxes became Local Education Tax. On the surface, it makes the local tax revenue look bigger, but it didn't make any change of substance since local governments only play the role of collecting Local Education Tax. 2. Tax revenue The total tax revenue in Korea was about 152 trillion Won in 2004, 8 of which 117.8 trillion Won being national tax revenue, and 34.2 trillion Won local tax revenue. The tax burden as a ratio of GDP, with social security excluded, was about 19.5% as of 2004. If the social security contributions are included, the total tax burden was about 25.0% as of 2004. 9 Korea expanded the social security system in the 1990s and as a result, as can be seen in Figure1, the tax burden rose rapidly from below 20% in early 1990s to around 25%. Without the social security contributions, the tax burden was about 18% in the early 1990s, and it increased by about two percentage points since then. Compared to the tax burden with social security included, the tax burden without the social security has been relatively stable for the past decade. 5
Special districts are the local governments below the metropolitan level, and cities/towns are the local governments below the provincial level. 6 These are Acquisition Tax, Registration Tax, Leisure Tax, License Tax, Community Facility Tax, Regional Development Tax, Local Education Tax, Inhabitant Tax, Automobile Tax, Farmland Tax, Butchery Tax, Tobacco Consumption Tax and Motor Fuel Tax. 7 These are License Tax, Property tax, Business Place Tax. 8 It is about 150 billion dollar as the exchange rate of one US dollar is about 1,000 Won. 9 It is an estimate since the social security contributions are not officially published yet. As of 2003, the tax burden with and without social security contributions was 20.4% and 25.4%, respectively.
When compared to the tax burden of other countries, Korea ranks second to the lowest (after Mexico) as can be seen in Figure 7. Other countries that show low levels of the tax burdens are United States, Japan, Ireland and Turkey. Except these countries, the tax burdens are above 30% and some countries have the tax burden as high as 50%. Therefore, the tax burden of Korea is fairly low by international standard. Since the re-birth of decentralization after more than 30 years of centralized era, the role of local taxes have been subject to debates, and, among the many potentially interesting topics, the debates are usually centered around the share of local tax revenue as a percentage of the total tax revenue. 10 Obviously, there are two ways to increase the role of the local taxes. The first is the rate increase of local taxes, and the other is the transfer of tax resources from the central to local governments. Interestingly, the debates about the way to strengthening the role of local taxes is concentrated on the latter. As can be seen from Figure 2, the share has been about 20% for the last decade, and it recently rose above 20% after a part of Education Tax became a local tax. Although not explicitly acknowledged by the ministries involved in this change, the purpose seems to be to increase the share of local tax revenue. However, as was discussed previously, the Local Education Tax is not really a local tax, and therefore the meaningful share of local tax revenue is still around 20%, which is slightly lower than the levels before 1995, the year when local autonomy was introduced. The major tax revenue sources in the national tax are VAT, Personal Income Tax (PIT), and Corporate Income Tax (CIT). The tax on fuels is an important revenue source in Korea and Transportation Tax ranks the fourth largest tax item. In Figure 3, the shares of the major tax items in the total national tax revenue are shown, 11 and the current share of VAT is 29.3%, 3 to 4 percentage points higher than the levels in the 1990s. The corporate income tax revenue shows a similar pattern, but the extent of the recent tax increase of CIT is greater than that of VAT. The share of CIT before 1998 was less than 15%, but it sharply rose to 19.2% in 1999, and it became 22.3% in 2003 and 20.9% in 2004. The tax revenue of PIT shows an apposite pattern. The share of PIT was around
This seems to be influenced by Japan, since Japan is probably the only other developed country in which the share of local tax is so importantly discussed as a measurement of successful decentralization. In most OECD countries, the debate on the tax assignment issue is centered around the citizens’ responsiveness toward local tax system. 11 Social security contributions are excluded in the total tax revenue in Figure 3.
22% in the middle of 1990s, and it reached 25.3% in 1998.12 Since then the share of PIT decreased below 20%, and its current level is lower than that of CIT. The tax revenue of Transportation Tax increased rapidly in the 1990s partly because the fuel tax rates are among the highest in Korea, and partly because the car ownership increased very fast in the 1990s. After late 1990s, the share of Transportation Tax is not increasing above 10%, because the increase in car ownership is getting slow. However, the share of fuel tax in Korea is still very high compared to other countries, and combined with VAT, it contributes to high share of indirect taxes in Korea. Table 1 shows the share of major taxes in the total tax revenue of selected countries. 13 The share of PIT in Korea is lowest with 12.7%, followed by France in which the share of PIT is 17.4%. Social security contributions is an important component of tax revenues in other countries, and the tax on personal income, combined with the social security, is 56.1 in France, by far higher than in Korea, where the share is 32.2%. The situation is similar for other countries, with the share of PIT and social security in Germany, Japan, U.K, and U.S being 54.4%, 56.7%, 46.9%, and 62.0%, respectively. In Table 1, the share of property tax in Korea is second to the highest with 11.8%. However, among the countries which collect a large amount of property tax, the situation in Korea is rather unique since much of the property tax revenue comes from so-called property transaction taxes such as Acquisition Tax and Registration Tax, which are local taxes. As a result, the major revenue sources in the local taxes are property transactions taxes. As is shown in Figure 4, the share of these taxes in the local tax revenue was more than 45% in the early 2000s, and it has come down to about 40% as of 2004. The recurrent property tax, which is the typical revenue source of local governments in other countries, is much less important in Korea with its share being about 10%. Other important revenue source for local governments is the Residence Tax, which is 10 percent surtax on PIT revenue. Its share was less than 10% in the early 1990s, but it has steadily increased to about 18% of local tax revenue. 3. Tax rates
The share of PIT peaked in 1998 because of the extremely high interest rates during the financial crisis in Korea.
The tax rates of PIT are in four brackets: 8%, 17%, 26%, and 35%. With surtaxes and social insurances added, the total tax wedges on employee earnings are 15.8%, 25.7%, 35.6%, and 45.5%. 14 Although the share of PIT in Korea is among the lowest in OECD countries, the nominal income tax rates in Korea are not particularly low. For example, the income tax rates in UK are 10%, 22%, and 40%. In US, they are 10%, 15%, 26%, 29%, 34%, 37.6% as of 2004. 15 What distinguishes Korea from other countries in terms of income tax burden is the width of income tax brackets. While average income earners in many OECD countries pay high marginal and average tax rates, those in Korea pays relatively low income tax rates. In table 2, the marginal and average income tax rates, with social security contributions included, are shown for different wage levels. In many countries such as Japan, Italy, Netherlands, Canada, UK, France, and Germany, the marginal tax rates paid by average wage earners (100% of APW) are the same or higher than those paid by higher wage earners (130% of APW). 16 On the other hand, the marginal tax rates paid by average wage earners in Korea is 5 percentage points lower than that paid by higher wage earners. In US, Austria, Switzerland and Nordic countries such as Finland, Norway, and Sweden, the marginal tax rates paid by average wage earners is much higher than that paid by higher wage earners. However, in these countries, the absolute levels of income tax rates are higher than Korea. When we compare the average income tax burdens, we reach similar conclusions. Overall, by international comparison, the income tax burden and tax rates in Korea are at very low levels. The tax rates of CIT in Korea have two brackets, and they have been cut several times over the past decade. As is shown in Figure 5, the high CIT rate was above 35% in the early 1990s, but it was decreased to 30.8% in 1997. Since then the CIT rates were cut twice in the 2000s, reaching the current level of 27.5% and 14.3%. 17 13
Social security contributions are included in the total tax revenue in Table 1. With the Residence Tax included, the total income tax rates are 8.8%, 18.7%, 28.6%, and 38.5%. The rates of social insurances for pension, health, and unemployment are, respectively, 2.15%, 4.5%, and 0.45%. 15 Before the introduction of EGTRRA, the income tax rates were 15%, 28%, 31%, 36%, 39.6%. To make an exact comparison, the local income taxes should be included in the US case. However, even with local income taxes included, the numbers are comparable to those in Korea. 16 In Japan, Italy, Netherlands, Canada and UK, the marginal tax rates for the average wage earners are the same as for the higher wage earners. In France and Germany, the former is higher than the latter. 17 The high CIT rate is 25% and the low CIT rate is 13%. As with PIT, Residence Tax is levied on CIT as 10% surtax. 14
- 10 -
In contrast to PIT, the share of CIT as a percentage of tax revenue is the highest in Korea among the selected countries. This is a curious feature since the corporate tax rate, with local tax included, is the lowest in Korea among these countries. 18 Since the tax revenue in Korea is much lower than these countries, the share of CIT as a percentage of GDP might show a different picture. However, Figure 6 shows that the share of CIT as a percentage of GDP is still the highest in Korea. The reason why Korea collects large amount of corporate tax with relatively low corporate tax rates is not thoroughly investigated yet. Intuitively, however, there may be several factors that cause this difference. Firstly, the deductions, exemptions and credits in the corporate tax code in other countries might be more generous than in Korea. Secondly, the financing structure of companies is different, and recently the debt levels of Korean companies have been significantly reduced. 19 Other possibility is income shifting. Perhaps there might be less incentives in the tax system in Korea than in other countries to shift income from corporate to share holder tax bases. 20 The rate of VAT in Korea is 10 percent flat. This is in contrast with most European countries, where the VAT rates are usually two or three brackets. The 10% VAT rate is relatively low compared to other OECD countries. Except Japan and Swiss 21 , the standard VAT rates are above 15%, and many countries have standard VAT rates higher than 20%. However, the low VAT rates in these countries are often below 5%. Although it is not clear from rate structures whether the 10% flat VAT rate in Korea is low or high by international standard, the share of VAT as a percentage of GDP shows that it is relatively low in Korea: in most European countries, the share of VAT as a percentage of GDP is 5 to 10 percent while that of Korea is about 4.6%. 18
In 2002, the corporate tax rates in Japan, France, Germany, U.K. and U.S. were, respectively, 42%, 34.33%, 38.36%, 30%, and 40% (approximate) (KPMG, 2002). 19 In 1995, the debt levels of Korean companies were much higher than now and the share of CIT as a percentage of GDP was 2.4%, which was lower than Japan (4.2%), UK (2.8%) and US (2.9%). However, the debt levels can not be the single reason for the difference since the corporate tax rates in these countries have cut down recently (Korea’s corporate tax rate has been reduced by 1 percentage point during the period). Also the shares of CIT as a percentage of GDP in France and Germany in 1995 were still very low with 2.1% and 1.1%, respectively. 20 Most of the corporate income taxes in Korea are paid a few large stock companies. As of 2003, the number of corporations were 303,462, and 88% of the CIT was paid by top 5% (15,482 companies) whose asset values were greater than 10 billion dollars. Thus the CIT paid by sole proprietorships, partnerships and limited liability is very small in Korea. In the US, C corporations accounted for 58 percent of all business income in 1999 (Slemrod and Bakija, 2004, p. 46). 21 The VAT rates in Japan and Swiss are, respectively, 5% and 7.5%.
- 11 -
IV. Tax Reform Issues 1. Economic growth Among several tax reform issues that are currently under debates in Korea, the issues of economic growth and unemployment perhaps draw the most attention. This is because the recent levels of the economic growth and employment in Korea are both low compared to historical trends. The concern over the role of tax policy with regard to the economic performance is of course not specific to Korea, and has lead to the recent tax reforms in many other countries. Germany, whose economic performance for the past decade has been rather low, has cut the corporate tax rates from 51.6% in 2000 to 38.29% in 2004. Japan, another industrialized country whose recent economic performance has been weak, has also cut corporate tax rates from 48% in 1999 to 42% in 2004. Even a country like United States which has enjoyed relatively high economic growth in the past decade has cut a significant amount of income tax, and the rationale for this tax reform is the motivation for better economic performance. 22 As the examples of these countries show, however, the impact of tax cut on economic growth is not obvious, and cutting tax rates often results in higher government debt rather than higher economic growth rate. There might be several reasons for this. Firstly, if cutting corporate tax rates is simply a response to the tax competition, the effect of lower tax rates on attracting economic activities such as capital investment will be limited. Secondly, the efficiency of the tax system depends not only on the level of the nominal tax rates, but also on the broad tax base, and this is not often achieved by tax reform. A third reason why it is not easy to observe a clear link between lower tax rates and economic performance might be that the benefit of lower tax rates materializes in the long run. Many theoretical and empirical studies on this issue also indicate that there is not a clear-cut relationship between lower tax rates and economic performance. In a comprehensive survey on this issue, Myles (2000) concludes that the effect of taxation on economic growth, if there is any at all, is relatively minor. 23 He also suggests that 22
Glenn Hubbard (). However, a recent study by Lee and Gordon (2005) finds that the corporate tax rate is significantly negatively correlated with economic growth while the effects of personal income tax rates are less clear. 23
- 12 -
there is no need to be overly concerned about growth effects when tax reforms are being planned. He however does note that the Ramsey rule principles apply both in the static economy and exogenous growth models, and human capital and R&D are the major factors that generate growth. The experiences of foreign countries and research results in the literature provide general lessons on the role of tax policy with regard to the economic growth. Since Korea, a small open economy, is faced with ever growing possibility of capital tax competition, maintaining the relatively low level of the corporate tax rates and capital income tax rates might be necessary for good economic performance. However, it is not clear whether a drastic cut of the corporate tax rates, not directly related to tax competition, is conducive to economic growth, since the resulting revenue loss leads to the reduction of investment in human capital and R&D that are critical components for economic growth. In this regard, maintaining revenue neutrality by cutting corporate tax rates along with base broadening, i.e., the reduction of the tax exemptions and tax credits, will be more appropriate. 2. Tax competition and FDI As for the tax competition issue, Figure 5 shows that the corporate tax rates (local tax included) in Korea has steadily decreased from 37.4% in 1990 to 27.5% in 2005. Even though many countries have also reduced their corporate tax rates to a significant degree, the corporate tax rate in Korea is still relatively low, although the gap is narrowing. China and Japan, the two countries with which Korea might be competing for capital investment, have corporate tax rates of 33% and 42%, respectively. These comparisons, however, might not be relevant in the discussion of tax competition since the tax treatment of FDI is in general different from domestic investment. For example, the overall comparison of preferential tax rates toward FDI indicates that China is more aggressive in attracting FDI with preferential tax treatment, although there is a sign that it might change in the future. There are of course vast literature on the desirability of preferential tax treatment of FDI as well as on the responsiveness of FDI to the preferential tax treatment. As for the desirability of preferential tax treatment of FDI, Janeba and Peters (1999) show that a
- 13 -
restriction on preferential tax treatment, by international cooperation such as OECD’s harmful tax competition guideline, results in higher tax revenue. However, Wilson (2004) and Keen (2001) show that the tax revenue is higher when preferential tax treatment of FDI is allowed. As Janeba and Smart (2003) discuss, one can obtain different results on this issue depending on different assumptions about tax competition environment. These studies are at an earlier stage, however, and there are few empirical findings. Therefore, it would be difficult at this stage to make a normative conclusion about the desirability of the restriction on preferential tax treatment of FDI. From a practical policy point of view, it seems unavoidable for Korea to offer to FDI preferential tax treatment comparable to other neighboring countries, since it is a common strategy for most of the Asian countries. 3. Redistribution It can be argued that, for the past thirty years, redistribution has been a less important issue in Korea than economic growth. This is partly because the high level of economic growth has generally increased the welfare of the most people, and partly because the pre-tax income distribution has been fairly even compared to many industrialized countries. All these situations are changing. Economic growth is not as fast as it used to be, and the benefit of it is not as widely spread out as it used to be. As a result the role of tax polices in the area of redistribution is recently gaining some attention. An introduction of EITC type income tax credit is currently being considered, and the tax burden of property tax has increased for the past two years after almost a decade of no change. The changes in tax policies with regard to redistribution policy are, however, limited to these episodes, and the tax burdens of some other tax items have actually decreased rather than increased. Both corporate and income tax rates have decreased, and special excise taxes on “luxury goods and services” have also cut down. 24 An exception is the tax on tobacco and its tax burden has recently increased almost by 40%.25 But obviously the increased tax burden on tobacco is not for redistribution purpose.
Luxury goods and services is legal definition of special excise tax and included refrigerator, golf course fee, etc.
- 14 -
When we examine the overall tax burden as a ratio of GDP, it shows that it has steadily increased from 20% in 1995 to 24% in 2004. Thus, the overall tax burden has increased somewhat, but it is due the effect of income elasticity of tax revenue, not as a result of nominal tax rate increases. Also, its level is still among the lowest in OECD countries, implying that the tax policy is not an active tool for redistribution policy in Korea. The share of indirect taxes such as VAT and excise taxes in Korea is about 55% of the total tax revenue. Since the overall tax burden is very low and the share of the indirect tax revenue is high, one can predict that widening income gap and higher unemployment rate will lead to higher tax burden and higher share of income tax in order to strengthen the redistributive function of tax polices. This is at least the implication one gets from international comparison of tax structure of OECD countries. As Figure 7 shows, the share of the income tax in the tax revenue in Korea is the lowest in OECD countries except Mexico. However, the implications from the theoretical discussions on the role of tax policies for redistribution are no more clear-cut than for economic growth. As for the issue of direct-indirect tax mix, the famous Atkinson and Stiglitz (1976) theorem indicates that indirect taxation is not necessary for welfare maximization. However as Boadway and Pestieau (2003) notes, there are circumstances under which the role of income tax is rather limited. Among the many reasons for the violation of Atkinson and Stiglitz theorem, the most notable reason in Korean tax policy seems to be the administrative costs of income taxation. When we hear public discussions on the role of income tax, the most frequently cited problem with income taxation in Korea is the horizontal inequity between employees and self-employed. Although the level of income tax burden of employees is by no means high by international standard, many employees complain that they are paying too much income tax compared to self-employed. The tax evasion by self-employed is actually a serious tax administration problem in Korea, and led to an introduction of a type of income tax deduction that is not found in other countries: as a means of making transparent the reported income of self-employed, employees are allowed to claim income tax deduction by a certain amount of expenditures under the condition that they are paid with credit cards. Although this 25
- 15 -
scheme contributed to mitigating horizontal inequity between employees and self-employed, 26 it also contributed to lowering the progressivity of income tax burden. 27 Another intended effect of this scheme was the broadening of VAT tax base, which, although not widely recognized by general public, implies a higher VAT tax burden for consumers. 28 Interestingly, this scheme turned out to be politically popular in Korea, despite the trade-off between vertical equity and horizontal equity. This episode is interesting in itself, but it seems that it can be also explained by the theoretical discussions on the direct-indirect tax mix. For example, Boadway et al. (1994) show that the role of indirect tax increases when income tax can be evaded. Although they do not deal with the cases where the degree of income tax evasion is different between employees and self-employed, the general implication one gets from this research is that the tax mix can be affected by tax evasion considerations. In a recent paper on this issue, Bird and Zolt (2005) provide more detailed description about the limited role of income tax for redistribution in developing countries. There are three reasons they put forth this proposition. First, in many developing countries, income taxes are virtually withholding taxes on labor income in the formal sector. Second, the costs associated with personal income tax system likely exceed the costs associated with other taxes. Third, given the ineffectiveness of personal income tax, the government may have to rely on indirect taxation to address the issues of poverty and inequality with redistributive expenditure programs. Related to the issue of direct-indirect tax mix is the role of expenditure policies for redistribution. Noting that the role of income tax is rather limited for redistribution purpose in developing countries, Bird and Zolt argue that expenditure policy, rather than tax policy, can be more effective tool for redistribution. This implies that the direct-indirect tax mix is not a critical criterion in determining the effectiveness of redistribution policy. In addressing this issue, however, Zee (2004) argues that, in 26 Since the amounts of sales transacted with credit cards are electronically sent to tax authorities, this scheme turned out to be quite effective in broadening the income tax base of self-employed. 27 There are two reasons for this. First, higher marginal rates are applied to the income deduction of high income earners. Second, about 40 percent of employees do not pay income tax in Korea. 28 The VAT incidence of this scheme depends on the relative magnitudes of supply and demand elasticities. As an example, the sector that is most noticeably influenced by this scheme is restaurant business, where most of transactions these days take place with credit cards rather than cashes. If the supply of restaurant is completely elastic, the burden will fall on consumers, perhaps in the form of lower quality of foods or higher prices.
- 16 -
designing equity-oriented expenditure programs, the role of progressive taxation is still important. However, the reason why Zee and Bird and Zolt reach different conclusions is because they are addressing different issues. While Bird and Zolt consider the problems of tax policy that arise from informal sectors and income tax evasion in developing countries, Zee considers the optimal income tax progressivity that is associated with income inequality ignoring the issue of income tax evasion. Since the recent discussions in the literature suggest that income tax evasion can limit the role of income tax, and expenditure policy can be more effective than tax policy for redistribution, the current structure of tax system in Korea does not necessarily imply that its redistributive function will remain weak. That is, we can predict that, as the tax administration becomes more effective, the role of income tax will gradually become more important. Also, if necessary, the redistributive function of the government can be strengthened by increasing the social expenditures. There can be, however, other explanations why redistributive function of the government in Korea is weak. The tax administration in Korea is actually not as inefficient as in developing countries since every citizen in Korea can be identified with identification number, and information technology in Korea is quite at high level. Therefore, the fundamental reason why the role of income tax in Korea is limited may be because tax policy is not a major political issue in Korea. Although there have been hot debates about soaring real estate prices and associated tax burden, 29 there have never been such intense attention given to income tax burden. This is unlike other countries such as United States where income tax burden have been cut down by 3%p of GDP in three years, and Nordic countries where dual income tax systems were introduced in the 1990s. 4. Political economy The political economy aspect of tax policies in Korea has not been subject to much academic debates yet, but a couple of hypotheses can be suggested. The most obvious reason why tax policy in Korea is not a major political issue is because political parties 29
The rapidly rising housing prices have been one of the major political issues in Korea. However, even with real estates, public opinion does not favor overall increase of the tax burden. This will be discussed in more detail in the later section.
- 17 -
have not yet been based upon their positions toward fiscal policies. It is notoriously known in Korean politics that political parities are divided by their affiliation with specific regions, notably south-east and south-west regions. This basically makes it impossible for political parties to establish a coherent position toward fiscal policies since their supporters include wide political spectrum from the unemployed to high income earners. The second reason why the debate about fiscal policy is inactive in Korea may be due to the fact that citizens in Korea have no experiences of actively participating in local government’s decision making. As a result, most of them do not differentiate between the local taxes and central taxes even after 10 years of decentralization. Thus decentralization in Korea has not yet contributed to making citizens to be aware about the impact of local government’s fiscal policies on their welfare. In summary, considering the fact that the tax burden and the share of income tax in the tax revenue is almost lowest among OECD countries, there seems to be more room for redistributive role of tax policy in Korea. This will take place gradually, however, as the tax evasion of self-employed becomes more difficult and political parties are established based upon their positions toward fiscal policies. Improvement of local tax system, while not directly related to the redistribution tax policy, will enhance citizens’ awareness of fiscal policies in general and make the public sector to function more efficiently. 5. Tax assignment The issue of tax assignment is quite controversial in Korea. It is often argued that local governments in Korea do not have enough tax resources, and as a result, the share of local taxes as a percentage of total tax revenue is as low as 20%. This is a curious argument, though, since the average of the share of local taxes in OECD unitary countries is about 13.8%. 30 When we examine the structure of local taxes, local governments have an access to income tax (both personal and corporate) as well as taxes on real estates. Since the revenue from Residence Tax (10 percent surtax on income taxes) as of 2004 was about 5 trillion Won, raising its rate from 10% to 15%
OECD Revenue Statistics (2004, p. 212)
- 18 -
will increase local tax revenue by 2.5 trillion Won, about 7% of local tax revenue. 31 An increase in income taxes by this amount is of course politically difficult, and no such discussions were brought up. In fact, as far as increases in local tax rates are concerned, no meaningful debates have ever occurred since the start of decentralization. Instead, the discussions on local taxes are usually focused on the transfer of tax resources form the central to local governments. Since local governments have an access to income taxes and property taxes, the tax resource currently under debates as a potential candidate for transfer to local governments is VAT. However, making a part of VAT a local tax in a small unitary country like Korea is not a desirable choice, and such a scheme mainly enriches the fiscal capacities of the local governments located in Seoul metropolitan region, which are least in need of transfer of tax resources. Those who understand these constraints but still want the transfer of tax resources basically argue a type tax sharing, which was recently introduced in Japan. 32 Even in the case of Japan, however, the introduction of local consumption tax was accompanied by an increase in VAT rate. 33 In understanding the aforementioned debates about local taxes in Korea, it should be noted that local taxes were in existence in Korea even before decentralization started in 1995. During the centralization era, local taxes were simply a means of transferring fiscal resources to local governments. Neither local governments nor local residents regarded local taxes as separate from national taxes. What distinguished local taxes from general grants, during that time, was the simplicity and transparency of local taxes as a means of fiscal transfers. This naïve concept of local tax is supposed to change with decentralization. A necessary condition for such a change to take place is that local residents perceive the link between the local tax burden and local public services. This has not happened yet for two reasons. The first reason is historical. Most citizens and politicians are still used to the concept that local taxes are simply a means of fiscal transfers from the central to 31
Local Tax Act allows local governments to raise Residence Tax rates above 50% of the standard rate. Japan introduced a local consumption tax in 1997. The local tax rate is determined by law, and the 20 percent of VAT collected by the central governments is distributed to provincial level of local governments based upon statistical data on consumption in each provinces. The provinces then distribute half of the local consumption tax to cities/towns based upon population and number of employees. 33 In 997, VAT rate in Japan were increased from 3 percent to 5 percent. 32
- 19 -
local governments. The second reason is the blurred division of expenditure responsibilities between the central and local governments. When local public expenditures are not clearly known to residents, it is natural that they desire that such services be provided by national taxes rather than by their own tax burden. Since the transfer of tax resources to local governments have been an important political issue, other aspects about tax assignment has not been much discussed. From a theoretical point of view, the most notable feature of local taxes in Korea is the concentration of tax resources around Seoul metropolitan area. Figure 8 shows per capita tax revenue of upper-level local governments, and that of Seoul is almost twice as much as the average. Since the local tax rates in Korea are the same for all local governments, this raises two possibilities. The first reason is that the residences in Seoul are on average much richer than those in other local governments. While it is to some extent true, it does not fully explain the reason since the difference of income levels between local governments is not so large, and the revenue of many local taxes is not elastic to income. The second reason is tax exporting to non-residents, mainly by taxing corporations, which are concentrated in Seoul metropolitan region. If a part of the high per capita tax revenue in Seoul results from tax exporting, its effect on local public finance is not negligible since tax exporting of 0.2 or 0.3 million Won per capita means 2 to 3 trillion Won of tax revenue. If the structure of local expenditures is such that per capita expenditure in large cities is high, the fiscal benefit from such large amount of tax revenue will be limited. However, the congestion effect of local expenditures in Korea is not high. 34 Therefore the fiscal benefit from the large amount of local expenditures in a big city like Seoul attracts fiscally induced migration into it. There are of course other reasons that make Seoul capital region so large, 35 but the fiscally induced migration certainly seems to play an important role in making Seoul metropolitan region among the largest in the world. 36 34 Figure 9 shows the per capita expenditure of upper-level governments. It is around 1 to 2 million Won for large cities (Seoul, Busan, Ulsan, GJ, DJ, IC, and DG) and Kyonggi province. It’s between 2 to 5 million Won for all other provinces. The congestion parameter of local public services in Korea is estimated to be around 0.55 (Kim, 2004) while the number is typically above 1 for other countries (Reiter and Weichenrieder, 1997). 35 Ades and Glaeser (1995) find that the size of capital region is related to protectionism, poor infrastructure, and political system such as dictatorship. 36 The population share of Seoul capital region is about 48%. The population shares of the metropolitan areas of Tokyo, London, Paris, Mexico City, and Kuala Lumpur are, respectively, 27%, 20%, 17%, 21%,
- 20 -
As are discussed in Boadway (2004) and Boadway and Hobson (1993), when there are different levels of net fiscal benefits accruing to otherwise identical individuals, fiscal equity requires that the differences in net fiscal benefits be equalized across local governments. 37 Also, from output-maximization point of view, fiscally induced migration causes inefficient allocation of labor across local governments. 38 More over, the related inefficiency cost may be very high in the case of Korea, since a large amount of fiscal resources are devoted to balanced regional development policy. 39 Given the fact that the design of tax and expenditure assignments can have a significant impact on agglomeration in Seoul metropolitan region, the current focus on local consumption tax in the debates about tax assignment seems too narrow. 40 Rather than focusing on a simple figure such as the share of local tax as a percentage of total tax revenue, the issue should be discussed with broader perspectives. 41 The first criterion that seems to be important in the discussion about tax assignment in Korea is the incentives faced by central and local governments. Since local governments are not interested in redistribution, the progressive parts of local taxes, especially Residence Tax and real estate taxes seem better suited for national taxes. 42 As for the recurrent property taxes on real estate, such a change was made in 2004. 43 As for Residence Tax, it is piggy-backed, with a flat rate, on the income tax revenue. A better scheme will be to make Residence Tax piggy-backed on income tax base, rather than income tax revenue. Since the income tax payment is concentrated in high income groups, this will make the distribution of local income tax more evenly distributed as well as leave room for expanding the role of local income tax. 44 and 17% (http://www.citypopulation.de/). From welfare-maximization point of view, migration induced by the economies of scales in the consumption of local public goods is not inefficient. 38 When migration is induced by differences in net fiscal benefits, the marginal productivities of labor are not equalized across regions. 39 Agglomeration in Seoul Metropolitan region gives rise to many regional policies that are often unsuccessful and inefficient. 40 When a part of VAT becomes local tax revenue based on a certain statistical criteria, it is not really a local consumption tax. But that is the terminology used both in Japan and Korea. 41 There is vast literature on this issue. See, for example, Bird (2000) and McLure (2001). 42 Nordic countries such as Sweden, Norway, and Finland introduced dual income tax system, and local governments apply flat income tax rates on the income tax bases determined by the central government. 43 Detailed discussion is in the next section. 44 This is the local income tax system adopted in Nordic countries. See OECD (1999) and Rattso (2005). 37
- 21 -
A second criterion that needs to be taken into account in the design of tax assignment in Korea is the issue of tax exporting. A significant part of local taxes is collected from companies, which are concentrated around Seoul metropolitan region. As was discussed previously, this allows local governments in the region to provide quality public services without making the residents take extra tax burden. This in turn induces agglomeration in the Seoul region and market-intervening regional policies. A related, and probably the most important criterion in the design of tax assignment is its link with expenditure assignment. In many developed countries, per capita expenditure in large cities is not less than that in smaller local governments. In Korea, even the subway constructions in large cities are subsidized by the central government. Education, the typical local public services in other countries, is also provided by the central government. Therefore, the current focus on transfer of tax resources to local governments should be shifted to the issue of the balance between fiscal capacities and expenditure responsibilities of local governments, especially large cities. 5. Taxes on real estates [To be completed]
IV. Conclusion [To be completed]
- 22 -
References Ades, A. and E. Glaeser, 1995, “Trade and Circuses: Explaining Urban Giants,” Quarterly Journal of Economics 109, 195-228. Anthony A. and J. Stiglitz, 1976, “The Design of Tax Structure: Direct versus Indirect Taxation,” Journal of Public Economics 6, 55-75. Bird, R., 2000, “Rethinking Tax Assignment: A Need for Better Subnational Taxes,” Tax Notes International 20, 2069-96. Bird, R., and E. Zolt, 2005, “Redistribution via Taxation: The Limited Role of Personal Income Tax in Developing Countries,” University of Toronto Working Paper. Boadway, R. and Flatters, F., 1982, “Efficiency and Equalization Payments in a Federal System of Government: A Synthesis and Extention of Recent Results,” Journal of Canadian Economics 15, 613-33. Boadway, R. and Hobson, P., 1993, Intergovernmental Fiscal Relations in Canada, Toronto: Canadian Tax Foundation. Boadway, R., M. Marchand and P. Pestieau, 1994, “Towards a Theory of the Direct-Indirect Tax Mix,” Journal of Public Economics 55, 271-288. Boadway, R. and Pestieau, P., 2002, “Indirect Taxation and Redistribution: The Scope of the Atkinson-Stiglitz Theorem,” Working Paper. Boadway, R., 2004, “The Theory and Practice of Equalization,” CESifo Economic Studies 50, 211-254. Keen, M., 2001, “Preferential Regimes Can Make Tax Competition Less Harmful,” National Tax Journal 54, 757-62. Kim, J., 2003, “Welfare Effect of Property Transaction Tax” paper presented at 59th IIPF conference. Kim, J., 2004, “Expenditure Assignment in Korea: Does It Spur Regional Concentration?” paper presented at 60th IIPF conference. KPMG, 2002, “KPMG Corporate Tax Rate Survey for 2002” KPMG, 2004, “KPMG Corporate Tax Rate Survey for 2004” Janeba, E. and Peters, W., 1999, “Tax Evasion, Tax Competition and the Gains from Nondiscrimination: The Case of Interest Taxation in Europe,” Economic Journal 109, 93-101. Janeba, E. and Smart, M., 2003, “Is Targeted Tax Competition Less Harmful Than its Remedies,” International Tax and Public Finance 10, 259-280.
- 23 -
Lee, Y. and H. Gordon, 2005, “Tax structure and economic growth,” Journal of Public Economics 89, 1027-1043. McLure, C., 2001, “The Tax Assignment Problem: Ruminations on How Theory and Practice,” National Tax Journal 54, 339-64. Myles, D., 2000, “Taxation and Economic Growth,” Fiscal Studies 21, 141-168. OECD, 1999, Taxing Powers of State and Local Government. Rattso, J., 2005, “Local tax financing in the Nordic countries,” Yearbook for Nordic Tax Research. Reiter, M. and Weichenrieder, A., 1997, “Are Public Goods Public? A Critical Survey of the Demand Estimates for Local Public Services,” FinanzArchiv 54, 374-408. Sandmo, A., 2005, “Inequality and Redistribution: The Need for New Perspectives,” Unpublished paper. Slemrod, J. and Bakija, J., 2004, Taxing Ourselves: A Citizen’s Guide to the Debate over Taxes, Cambridge, Mass.: MIT Press. Wilson, D., 2004, “Tax Competition with and without Preferential Treatment of A Highly-Mobile Tax Base,” Georgia State University Working Paper. Zee, H., 2004, “Inequality and Optimal Redistributive Tax and Transfer Policies,” Public Finance Review 32, 359-381.
- 24 -
15 With Soc. Sec.
Without Soc. Sec.
Fig. 1. Tax revenue as a share of GDP
0.2 Without Local Education Tax 0.1
96 97 year
Fig. 2. The share of local tax revenue
- 25 -
Corporate Income Tax 0.1 Transportation Tax
01 02 03 04
Fig. 3. Tax shares in national tax revenue (1981-2004) Note: Social security contributions are excluded.
Table 1. Tax revenue of major taxes as % of total tax revenue (2002) Individual income tax
Corporate income tax
Goods and services
Note: Social security contributions are included. Source: OECD Revenue Statistics (2004)
- 26 -
Acquisition Tax and Registration Tax 0.4
Land Tax and Property Tax
Fig. 4. Tax shares in local tax revenue
- 27 -
Table 2. Tax wedges on labor income Marginal PIT and social security Average PIT and social security contribution rates on gross labor contribution rates on gross labor income income Country % of APW % of APW 67% 100% 133% 167% 67% 100% 133% 167% 35.4 35.4 51.4 51.4 25.2 28.6 32.4 36.2 Australia 59.6 55.5 61.1 61.1 38.6 44.9 47.7 50.4 Austria 74.1 66.4 69.8 70.2 46.9 54.2 57.8 60.3 Belgium 35.1 33.9 33.9 36.6 29.0 32.3 32.8 33.2 Canada 44.9 48.1 48.1 51.4 41.9 43.6 44.7 46.0 Czech Republic 42.9 49.2 62.9 62.9 39.4 41.5 46.7 49.9 Denmark 50.5 55.1 59.7 59.7 38.6 43.8 47.2 49.7 Finland 55.5 66.6 53.5 53.5 32.5 47.4 49.9 50.6 France 58.8 64.0 60.2 65.0 45.4 50.7 54.0 55.7 Germany 34.4 44.2 54.1 54.1 34.4 34.9 37.7 41.0 Greece 54.2 54.2 76.9 63.3 38.7 43.8 50.1 54.5 Hungary 40.4 40.4 40.4 45.2 24.4 29.7 32.4 38.7 Iceland 30.0 33.2 53.0 49.9 15.6 23.8 30.8 35.0 Ireland 52.7 58.0 58.0 59.1 41.7 45.7 48.5 50.5 Italy 27.8 31.5 31.5 39.6 25.8 26.6 27.8 29.3 Japan 17.8 24.8 29.7 29.7 15.3 16.6 19.5 21.5 Korea 36.5 45.9 53.8 53.8 27.5 31.9 36.3 39.8 Luxembourg 15.4 23.4 23.4 33.7 10.6 15.4 18.8 21.1 Mexico 55.6 50.7 50.7 52.0 38.1 43.6 39.7 40.5 Netherlands 21.0 33.0 33.0 39.0 19.0 20.7 23.8 26.5 New Zealand 43.2 43.2 55.1 55.1 33.8 36.9 40.4 43.4 Norway 45.7 45.7 45.7 45.7 41.9 43.1 43.8 44.1 Poland 37.8 39.4 47.5 47.5 29.6 32.6 35.9 38.2 Portugal 48.3 48.3 48.3 38.8 42.0 43.6 44.5 Slovak Republic 48.3 49.4 45.5 48.4 48.4 33.6 38.0 40.2 41.9 Spain 51.7 51.7 64.1 63.5 46.2 48.0 49.6 52.4 Sweden 30.8 36.5 40.6 41.1 26.1 28.8 31.1 33.0 Switzerland 44.5 44.5 48.0 48.0 41.8 42.7 43.5 44.4 Turkey United Kingdom 40.6 40.6 40.6 31.7 26.4 31.2 33.6 34.2 United States 34.1 34.1 43.4 43.4 27.3 29.6 32.6 34.8 Note: APW stands for average production wage. Source: OECD Taxing Wages, 2004.
- 28 -
35.2 33. 30.8
30 22. 19.8
Figure 5. Corporate Tax rates change in Korea (1990-2005). Note: Local tax included.
42 40 38.36
Figure 6. CIT rates and GDP share of CIT (2002). Note: Local tax included. Source: OECD Revenue Statistics, 2004. KPMG (2004).
- 29 -
Denmark Finland Norway
Australia Switzerland Ireland
40 Gini coefficient
Figure 7. Tax revenue and distribution of income after tax (1998). Source: World Income Inequity Data, OECD, Revenue Statistics (2004)
Table 1. Transition from aging to aged society of selected countries Japan
- 30 -
Jeju Ulsan DJ
IC Busan CN GN DG GJ GW GB CB JB JN
Figure 8. Per capita tax revenue of local governments
4 JN Million Won
JB GB CB
2 Kyonggi Ulsan
Figure 9. Per capita expenditure of local governments
- 31 -
35 Greece Korea
% of Self -employed
20 Czech 15
Ireland Poland UK Australia Japan G erm any Slovak Netherlands Finland Austria Canada Sweden US France Luxembourg Norway
30 35 Income tax burden
Figure 10. Income tax burden and self-employed Source: OECD Revenue Statistics (2004), OECD Labor Statistics (2004)
- 32 -