The Tax Models in Japan and Korea: Concepts and Evidence from a Comparative Perspective

Abstract Existing studies have relied on the notion of developmentalism to explain key aspects of the tax policies in Japan and Korea. However, limited efforts have been made to explore these cases from a comparative perspective based on relevant evidence. Far fewer studies have been conducted for examining the contemporary evolution of the tax policies following major reforms since the 1990s. This article seeks to fill these gaps in the research. Employing an analytic framework of tax structure, it provides key definitions of the old and new tax models in Japan and Korea in a way that is comparable with other OECD cases. “Residualism” and “constrained activism,” two heuristic models drawn from low-tax OECD countries, provide useful references for this comparative task. To validate key assessments, the author utilizes and replicates extensive tax data that operationalize important aspects of the tax structure from the 1980s to 2018.


Introduction
Until the 1990s, Japan and Korea had a unique tax model that was different from that of other OECD nations. Developmentalism, which underscored a government's commitment to investments and savings over public spending and consumption (Drucker 1993;Holliday 2000;Haggard 2018;Johnson 1982;Kwon 1997), provided a key template for exploring how tax policies in these countries were instituted. Total taxation was set at a distinctively low level. However, taxes were collected in a progressive way, targeting high wage earners, established corporations, and large asset holders. Other income-tax payers were awarded extensive tax expenditures in the form of exemptions and deductions. Within these common features, Japan presented higher taxation than Korea. Aggregate tax burdens were also distributed differently between direct taxes (wage and capital income taxes) and non-direct taxes (social security contributions (SSC) and consumption taxes), depending on their policy priorities (Dewit and Steinmo 2002;Ide and Steinmo 2009;Ishi 2001;Johnson 1982;Kim, D.-K. 2018;Kim, M.-K. 2018).
Existing studies have documented these tax features fairly comprehensively (Choi and Hyun 1997;Choi et al. 1991;Ishi 2001;Keen et al. 2011;Keen 2008;Kim, M.-K. 2018). However, limited efforts have been made to establish a conceptual and empirical framework to identify key aspects of tax policies in Japan and Korea from a broader OECD perspective. Far fewer studies have been conducted for examining the contemporary evolution of tax policies following major reforms since the 1990s. Beyond speculations on a system shift toward neoliberalism (Dewit and Steinmo 2002;Ide and Steinmo 2009;Kim, D.-K. 2018;Kim, M.-K. 2018), no comprehensive analysis has been provided yet.
The present study addresses these gaps in the research. It first defines the core features of the tax model in Japan and Korea in comparison with other alternative models, "residualism" and "constrained activism," that the author drew from low-tax OECD countries. Residualism builds upon a liberal market idea in which governments impose low taxation to ensure that private economic actors have much of their income and wealth at their disposal. However, the overall tax level is higher than in the developmentalism cases because the governments provide a range of targeted welfare benefits for disadvantaged citizens in the market. Under this framework, the governments have instituted a progressive method of tax collection that imposes heavy responsibilities on high-income groups and asset holders. They also rely more on direct than non-direct taxes, which have a broad base with tight restrictions on exemptions and deductions. Meanwhile, countries with constrained activism feature even more public spending than the residualism cases, although their overall tax level is nevertheless below the OECD average. Governments collect taxes either progressively or regressively, along with diverse patterns of tax base. Depending on public welfare programs and tax policy priorities, they present diverse sub-cases that combine progressive taxation with broad direct taxes, regressive taxation with non-direct taxes, and regressive taxation with broad direct taxes.
Following an evaluation of these comparative typologies, Japan and Korea provided ideal cases for developmentalism until the 1990s. While taxes were collected in a similar progressive way as in residualism cases, total taxation remained lower. Tax expenditures were also extensively available for income taxpayers. Since the 2000s, Japan and Korea have begun experiencing notable changes in their tax policies. Japan now belongs to the group of constrained activism, while Korea is making a transition toward fiscal residualism. In a departure from the past, both countries are moving toward more general OECD cases. A unique tax model that has dominated academic debates in East Asia no longer exists.
The present study examines these evolving features of the tax policy by employing an analytic framework of tax structure. It addresses major tax dimensions by attending to total taxation, progressivity, and the tax base (Beramendi and Rueda 2007;Cusack and Beramendi 2006;Prasad and Deng 2009). Multiple indicators and tax data are introduced to operationalize each of these dimensions from a comparative perspective. The study also presents the most recent data to trace the evolution of contemporary tax policies in Japan and Korea. Accordingly, the author imports relevant tax indicators from the OECD database and replicates various existing formulae to extend their coverage.
Existing studies provide a wide range of explanatory accounts for tax policies in Japan and Korea. They combine various structural causes-such as population aging, a decline in the traditional family structure, low economic growth, rising inequality, and financial globalization-with other social and political institutional determinants, including social pressure from business groups, asset holders, and broad consumer groups; plurality-based electoral formulae; partisan politics between conservative and progressive liberal parties; and the low-tax legacy from developmentalism (Park 2022;Chopel, Kuno, and Steinmo 2005, 28-29;Dewit and Steinmo 2002;Estévez-Abe 2008;Ishi 2001;Kwon 2014;Park 2015;Peng 2004;Kim, M.-K. 2018;Yang 2017). However, their accounts have been limited with their analytic focus either on a broad paradigm change (i.e., if a neoliberal shift occurred in Japan and Korea) or detailed policy outcomes for specific tax categories. Relying on the three analytic dimensions of the tax model, this study provides a comprehensive look at subjects that are not only multi-faceted but also include new features comparable to broad OECD cases. This warrants a fine-tuned causal account that goes beyond the East Asian context.
The subsequent sections elaborate on the concept of fiscal developmentalism in comparison with residualism and constrained activism. Major findings and challenges in the literature concerning tax policies in Japan and Korea are reviewed, along with the key conceptual and empirical claims this study presents. These claims are scrutinized against various tax data involving the two Asian and other relevant OECD cases. The concluding section summarizes the major findings of the study and discusses implications for future research.
Varieties of taxation in low-tax countries Scholars have explored various tax models in developed OECD countries by considering three broad dimensions of tax policy (Akgun, Cournède, and Fournier 2017;Beramendi and Rueda 2007;Cusack and Beramendi 2006;OECD 2017;Plümper, Troeger, and Winner 2009;Prasad and Deng 2009;Swank 2016). First, total taxation refers to the overall tax level that a government extracts from the economy. Second, progressivity refers to how a government distributes the tax responsibility across different income/wealth groups, which varies depending on the government's preference for redistribution. The tax base, the last dimension, assesses key features of aggregate tax burdens that support the government's spending policy. It highlights the pattern of tax burdens across major tax categories and, particularly for this study, between direct and non-direct taxes. It also illustrates the extent of tax expenditures available for taxpayers, mostly for direct income taxpayers, for whom exemptions and deductions are primarily designed. A comprehensive map of how governments tax citizens can be drawn by combining all these features. Japan and Korea have long been considered ideal cases for fiscal developmentalism. An emphasis on massive investments and savings over public spending and consumption-which ultimately spurred rapid, export-led economic growth-resulted in a tax system in which the tax level was the lowest of all industrialized countries. Public spending remained underdeveloped. Tax-based services and assistance were scarcely provided, with large segments of insurance risks (unemployment, sickness, and retirement) being covered by families and companies (Drucker 1993;Haggard 2018;Holliday 2000;Johnson 1982;Kwon 1997).
Under this policy framework, the governments developed a progressive tax scheme whereby high wage earners, established corporations, and asset holders made higher contributions than most other economic groups. War experiences as well as a broad policy consensus that the core beneficiaries of rapid economic growth should pay a major proportion of tax burdens provided a political justification for this tax scheme (Brownlee 2009;Ide and Steinmo 2009;Johnson 1982;Yun 1997, 99-121). When considering the tax base, the governments provided an extensive array of exemptions and deductions for direct income taxes, mostly targeting middle-to-low wage earners, companies, and asset holders (Brownlee and Ide 2017;Choi et al. 1991, 219-221, 285-291;Yang and Min 2013, 57-64). This preferential policy was deemed necessary to mobilize massive private savings and investments and also provided an effective measure to compensate for the lack of public welfare programs available for citizens. Within these common features, Japan collected higher taxes than Korea. Tax burdens were also distributed differently across major tax categories, depending on public welfare commitment and other historical contingent factors. Table 1 summarizes key features of the developmentalist tax model outlined so far. The subsequent section provides detailed discussions regarding within-group differences.
Meanwhile, several OECD countries have developed alternative models for low taxation. Here, conventional wisdom looks at a limited number of countries whose tax policies have been guided by the liberal market idea (Beramendi and Rueda 2007;Cusack and Beramendi 2006;Peters 1991;Prasad and Deng 2009). This study finds that more countries present low-tax cases by employing diverse tax features. These differences are conceptualized into two heuristic tax models: residualism and constrained activism.
Residualism is best illustrated by such countries as Australia (since the 2000s following significant tax cuts), Switzerland, and the US, where the conventional liberal market idea has been influential in tax policy making (Bloch et al. 2016, 29;Brownlee 2016;Davis et al. 2019;Steinmo 1993, 136-138, 142-144). Their tax policies mirror the logic of residualist public welfare, where governments provide only basic-level welfare programs, leaving citizens in charge of further material wellbeing. Governments nevertheless intervene actively with targeted benefits if citizens prove incapable of sustaining minimal living conditions from the market (Esping-Andersen 1990; Huber and Stephens 2001).
Applying this policy logic to public taxation, governments retain total taxes at a much lower level than the OECD average, so that taxpayers can use a major proportion of their income and wealth at their discretion. However, their tax level is nevertheless higher than in the developmentalism cases, because of their commitment to targeted benefits for the poor and other disadvantaged economic groups. Within this framework, governments have designed a progressive tax policy that favors middle-to-low economic groups at the expense of more advantaged ones who can afford to pay higher rates. Tax burdens also mirror the residualist welfare programs that feature slim benefits in both tax-based services/assistance and contribution-based insurances. Governments financed these programs mostly by relying on broad, direct income taxes that had already been established since the first half of the twentieth century (Peters 1991;Prasad and Deng 2009;Slemrod and Bakija 2017, 19-28;Steinmo 1993;Wilensky 2002). Compared to other high-tax cases in Europe, which instituted more generous insurance benefits and universal public services and assistance, the governments in residualism cases did not have to seek large sums of additional revenue from SSC and value-added tax (VAT) (Beramendi and Rueda 2007;Cusack and Beramendi 2006). Instead, they upheld their revenue base by limiting the range of exemptions and deductions for direct income taxpayers (Brownlee 2016, 240-241;Carling 2015;Feuntes 2013, 11, 17;Michelmore 2012, 142-148). 1 Meanwhile, other OECD countries have developed alternative tax policies that suit the notion of constrained activism. Here, governments seek higher taxes to support more generous social spending than in the residualism cases, although their tax commitment is nevertheless "constrained" because the overall tax level is below the OECD average and, particularly, the Continental and Northern European levels (Beramendi and Rueda 2007;Bloch et al. 2016, 29;Cusack and Beramendi 2006;Petmesidou 2019). To meet their spending commitment, the governments have instituted diverse tax schemes, progressive or regressive, along with various forms of tax base to extract more revenues from citizens.
Australia before tax cuts in the 2000s and Canada after tax cuts in the late 1990s represent a good example of constrained activism (Bibbee 2008;Campbell and Murray 2018;Fenna 2007, 339-346). They have maintained higher taxes to match their policy commitment toward more generous insurance programs, alongside basic public services and assistance (Bochel 2019, 180-181;Cox 2019, 201-204;Scruggs and Allan 2006, 64). To support these expanded welfare programs, the governments have developed a tax scheme that combines progressive taxation with broad direct taxes. Like in the residualism cases, they place high tax rates on wealthy populations with high incomes. However, the governments rely more heavily on direct taxes to finance their spending programs. They either impose higher burdens on direct taxes vis-à-vis non-direct taxes than in the residualism cases, or further restrict the scope for exemptions and deductions for direct income taxpayers. The UK exhibits another case of constrained activism that is supported by regressive taxation and broad direct taxes. Its tax system used to be similar to that of Canada and Australia, along with similar public spending priorities (Migali 2004, 272-275;Steinmo 1993, 144-154). Since Thatcher's tax reforms in the 1980s, however, the UK's tax policy has become more regressive along with the rise in VAT and cuts in income tax rates (Prasad 2006, 103-121). Nevertheless, the tax base is dominated by direct taxes owing to tight restrictions on exemptions and deductions, as well as high effective taxes on capital income and particularly property income and capital gains (Adam, Browne, and Heady 2010, 24-26;Migali 2004, 287;Tetlow and Marshall 2019, 28).
Southern European countries, such as Greece, Portugal, and Spain, present yet another case of constrained activism that combines regressive taxation with nondirect taxes. Governments in these countries have provided hefty public insurances at the expense of basic public services and assistance. Public pension programs, for instance, have boasted the most generous retirement plans in Europe even in comparison with Continental and Northern cases (Petmesidou 2019, 163-166). To support these policy commitments, governments have focused their attention on consumption taxes in addition to heavy SSC. Large informal sectors embedded in the economy have further promoted this financial solution by limiting the role of direct income taxes as reliable revenue sources, thereby leaving non-direct taxes a primary option to mitigate fiscal shortage. Predictably, governments have instituted a regressive tax system where consumption taxes are collected at higher rates than progressive income taxes, as standardized against the OECD average. Non-direct taxes also assume heavier burdens than direct income taxes which continue to provide ample room for exemptions and deductions for the incomes earned from informal economic activities (Brys, Matthews, and Owens 2011;De Cos and Rodríquez 2014).

Literature review and contributions
The tax model in Japan and Korea until the 1990s Scholars have built on the notion of fiscal developmentalism to explore key aspects of tax policies in Japan and Korea. While highlighting such common features as lowest taxation, progressive tax collection, and extensive tax expenditures (Brownlee and Ide 2017;Choi et al. 1991, 219-221, 285-291;Yang and Min 2013, 57-64), they have nonetheless identified considerable differences between the two cases.
First, total taxes were much lower in Korea. Facing serious challenges from the Communist North and the lack of political legitimacy under an authoritarian rule, the Korean government was eager to achieve faster economic growth. Ultra-low taxes were deemed necessary to promote this goal by tightly controlling labor and business costs in export markets (Kwon 2014;Yang and Min 2013). Meanwhile, the tax level was higher in Japan. The Japanese government needed more revenues to finance health and pension programs that had been steadily expanded throughout the post-war period (Campbell 1992;Estévez-Abe 2008). However, total taxes managed to be lower than in the residualism cases because the welfare benefits were not as generous as in other OECD countries, and a young population profile, along with a vibrant labor market in the post-war economy, helped the Japanese government keep the cost of public insurance programs at a much lower level than in other OECD countries (Park 2020).
Partly reflective of these differences, Japan and Korea developed diverse patterns in tax burdens across major tax categories. In Korea, the near absence of public services/ assistance and public insurance programs led to only marginal tax burdens on SSC as well as wage and capital incomes (Yang 2017). This allowed consumption tax-which was expanded with a low 10 percent VAT rate to mitigate revenue shortages-to become the largest contributor to public revenue (Ahn 1997, 274;Na 1997, 154, 180-182). The situation was different in Japan where the health and pension programs were supported by a combination of moderate SSC and high capital taxes. The ruling Liberal Democratic Party promoted this unique financing mechanism. Amid concerns of the rising SSC burdens on ordinary employees, employers, farmers, and the self-employed who constituted their main electoral clients, LDP leaders decided to increase corporate taxes for large companies to subsidize the costs of health and pension benefits. Their decision led to capital income becoming the largest contributor to public revenue, eventually making direct taxes bear more burdens than non-direct taxes (Ishi 2001, 6-12). This solution also helped consumption taxes continue to assume the lowest burdens by delaying the introduction of a general VAT until the end of the 1980s (Brownlee and Ide 2017;Ide and Steinmo 2009;Ishi 2001, 43). The upper section of Table 2 summarizes these conventional tax policies in Japan and Korea.
Existing accounts, cited above, help in understanding the developmentalist tax model in Japan and Korea, as well as their specific differences. However, their conceptual framework has not been fully elaborated since the idea of fiscal developmentalism has been examined without systematic comparison with other relevant models from OECD countries. It is still not clear in what aspects the East Asian countries were distinctive enough to defy broad cross-national comparison. The present study addresses this issue by explicitly comparing the Japanese and Korean cases with the two alternative models, residualism and constrained activism, from other low-tax countries. It also analyzes the newly emerging tax models in Japan and Korea, further extending the coverage of the existing scholarship on these countries.

New tax models since the 2000s
Since the 2000s, Japan and Korea have been witnessing significant changes to their tax policies. First, SSC was expanded in both countries, particularly in Japan. Several structural forces, such as an aging population, the decline of the traditional family structure, low economic growth, and rising inequality, have contributed to the expansion by increasing the social demand for pension and health programs while reducing the pool of contributors (Dalsgaard and Kawagoe 2000, 10;Chopel, Kuno, and Steinmo 2005, 28-29;Yang 2017). Important changes also occurred with regard to capital taxation. In Japan, both statutory rates and aggregate burdens were cut down (Dewit and Steinmo 2002), whereas in Korea the statutory rates decreased but the tax burden increased with reduction in various tax expenditures (Kim 2006;Park 2015). Considering that many other OECD countries have also introduced capital tax reforms (in particular, corporate tax reforms) over recent decades by lowering statutory rates while expanding the tax base, Korea fits adequately into this general reform trend. Japan, however, is a rather deviant case. Existing studies explain these differences by referring to the mediating role of domestic political institutions and partisan politics (Park 2022;Hallerberg and Basinger 1998;Swank 2016). Finally, in Japan, consumption tax burdens expanded moderately following the introduction of a 3 percent VAT in 1989, with a gradual increase up to 10 percent in 2019 (De Mooij and Saito 2014;Ishi 2001;Kamikawa 2016, 41-42;Nippon Times 2019).
Scholars have made efforts to understand the consequences of these changes. Emphasizing the regressive implications from capital tax cuts and consumption tax increases in Japan, they have made a sweeping projections toward neoliberal transformation of public taxation (Ide and Steinmo 2009;Dewit and Steinmo 2002). The debate has been more nuanced in Korea. While some have claimed a similar neoliberal shift (Kim, M.-K. 2018;Kim, D.-K. 2018), others have endorsed a more moderate view by emphasizing the Korean government's effort to expand its tax base (Kim 2006, 7-43;Park 2020). Beyond these broad speculations, however, no systematic analysis has been offered to examine core features of the newly emerging tax models in comparison with their past and other OECD cases.
The present study makes a comprehensive effort to address this gap in the research. All substantive claims made in this section have been confirmed with the evidence reported in the following empirical sections. Starting with Japan, the country has departed from the previous developmentalist tax model to become a new case of constrained activism. With a surge in public insurance spending to support citizens in the world's most aged society, along with the expansion of previously underdeveloped public services and assistance targeting young families and women in the labor market (Estévez-Abe 2008; Fleckenstein, Lee, and Choi 2021), Japan's tax level is higher than in other residualism cases. The country continues to retain progressive taxation. To support its extended spending commitment, however, it has started to develop a new pattern of tax burdens in which SSC bears the largest share of all tax categories, even surpassing capital taxes, whose contributions have been shrinking in the wake of corporate tax cuts since the 1990s. Meanwhile, wage income and consumption taxes (despite the staged increases in the VAT rate over the last 30 years) have continued their lower standing. Overall, Japan relies more on non-direct taxes than direct taxes which still allow for extensive room for exemptions and deductions.
Viewed from other cases of constrained activism, where governments support their enhanced spending commitment by combining progressive taxation with broad direct taxes (Australia and Canada), regressive taxation with broad direct taxes (the UK), or regressive taxation with non-direct taxes (Greece, Portugal, and Spain), Japan presents a novel pattern where the government combines progressive taxation with non-direct taxes. The guiding principle is that the government seeks to increase tax revenues while preserving its developmentalist legacy. A limited group of advantaged populations still contribute to a major share of the public budget, while most other economic groups are awarded an extensive array of exemptions and deductions for wage and capital incomes (Ide and Steinmo 2009;Dewit and Steinmo 2002). With this tendency prevalent in the tax policy, the Japanese government has determined non-direct taxes as more prominent resources to support their enhanced spending programs.
Korea has also experienced notable changes in its tax policy. Providing only slim welfare benefits, the Korean government has nonetheless expanded the public insurance coverage to include diverse groups of vulnerable citizens (Yang 2017). Public services and assistance programs, previously underdeveloped, have also extended their basic benefits to accommodate various welfare needs from families, the aging poor, women, and young people (Fleckenstein, Lee, and Choi 2021). All these changes have led overall taxation to approach the level of the residualism cases. While taxation continues to be progressive, the Korean government has also introduced important changes to its tax base. It is worth noting that SSC exhibited a notable increase to finance contribution-based insurance benefits, but capital taxes expanded further to support tax-based services and assistance. The government played an important role in this transformation by cutting back on exemptions and deductions on capital taxes that used to be widely available for firms, family businesses, the self-employed, financial income earners, and property owners (Ahn 1997, 283-305;Kim, H.-S 2012, 247, 249;Lee 2012, 513, 551-552). Consequently, capital income has begun to make the largest contribution to public revenue, even larger than the SSC, which has also steadily expanded. Consumption, which used to be the largest contributor to public revenue, has been ranked third with little change from the past, only being followed by the wage income taxes, which still retain an extensive array of exemptions and deductions. All combined, Korea relies more actively on direct than non-direct taxes to support its enhanced spending programs.
Overall, Korea is making a transition to the residualist tax model, sharing similar important features in the areas of total taxation and progressivity. Direct taxes, driven by capital income taxes, have also begun to play a more important role than nondirect taxes. It is true that the developmentalist legacy is still at play-as shown by the extensive room for tax expenditures for wage income taxes. Unlike Japan, however, Korea has managed to reduce various exemptions and deductions for capital income taxes, paving the way toward the residualist tax model. The lower section of Table 2 summarizes key features of the new tax models in Japan and Korea thus far discussed.

Measuring the tax structure
Existing studies have sought to operationalize tax policies in OECD countries by focusing on three major dimensions of the tax structure-total taxation, progressivity, and the tax base-and providing a wide range of rigorous indicators (Akgun, Cournède, and Fournier 2017;Beramendi and Rueda 2007;Cusack and Beramendi 2006;OECD 2017;Plümper, Troeger, and Winner 2009;Prasad and Deng 2009;Swank 2016). Japanese and Korean studies have employed a similar approach by examining various statutory, marginal, and effective tax data, covering income to consumption; see Dalsgaard and Kawagoe (2000), Steinmo (2009), Ishi (2001), and Keen et al. (2011) for Japan, and Choi and Hyun (1997), Kim, M.-K. (2018), and Kim, D.-K. (2018) for Korea. However, their measurement has not produced comprehensive indicators that link data to concepts. It remains unclear, for instance, how to operationalize the progressivity of taxation and the structure of the tax base; nor is it certain how to compare these data in a cross-national setting. This section outlines these challenges comprehensively, along with feasible solutions to them.
Among the three dimensions of the tax structure, there is limited debate as to how to measure total taxation. Scholars have merely considered all tax revenues as a percentage of total economy (Dalsgaard 2000, 50;Dewit and Steinmo 2002, 160;Ishi 2001, 6-7;Keen et al. 2011, 6;Kim, M.-K. 2018, 101;Lee 1991, 187;Yang and Min 2013, 59). In this study, the tax level is measured by calculating the general government revenue as a percentage of GDP. This figure is standardized against the OECD average to facilitate cross-national comparison. Using a standard technique of hierarchical cluster analysis with the average group linkage method (Everitt et al. 2011), sample countries have been evaluated as groups with similar tax levels.
Addressing other dimensions of the tax structure has been more complicated. Ideally, progressivity can be measured by considering all diverse tax payments across different economic groups (Piketty and Saez 2007;Prasad and Deng 2009;Saez and Zucman 2019). Given the difficulty of producing such extensive tax data across countries, Japanese and Korean studies have instead focused on certain progressive tax categories, such as personal and corporate income taxes, and analyzed their statutory or marginal rates across different income groups and countries (Choi et al. 1991 Dalsgaard 2000, 42, 51, 54;Ide and Steinmo 2009, 130;Kim 2009, 208;Kim, M.-K. 2018, 274). While this approach appears practical from a comparative perspective, its narrow focus on progressive taxes has resulted in an uncomprehensive assessment of the overall progressivity of taxation. Even if wage and corporate tax rates are high, the tax system would not be as progressive if regressive tax rates such as the VAT rate also remain high.
The present analysis addresses this challenge by considering both the progressive and regressive tax rates in an integrated framework. It standardizes the highest rates of personal and corporate income taxes, averages them, and compares this figure with the standardized VAT rate. If the former is greater than the latter, the tax system is considered progressive. If the latter is greater, the tax system is regressive. A natural way to pursue this task is to compare statutory income tax rates with the VAT rate. While providing a sound ground for a cross-national analysis, statutory data are often misleading because high-income groups can be relieved of tax burdens by means of various exemptions and deductions. This issue is addressed by looking at a marginal effective tax rate for high-income groups and comparing it with the VAT rate. The OECD database (https://stats.oecd.org/) provides a readily usable proxy: the marginal effective tax rate for one-earner and two-earner families with 167 percent of the average gross wage earnings. Based on these tax data, we can consider that a tax system is progressive if both the statutory and marginal effective income tax rates are greater than the VAT rate. If both the rates are lower than the VAT rate, the tax system is regressive. All other results are deemed inconclusive.
Studies have also devised various indicators for the tax base, gauging the way that a government distributes aggregate tax burdens to support spending commitment. Overall, the tax base varies depending on the pattern of tax burdens across major tax categories and, particularly for this study, the extent to which direct taxes (wage and capital income taxes) contribute to public revenue vis-à-vis non-direct taxes (SSC and consumption taxes). It also highlights the extent of exemptions and deductions available for income tax payers.
To measure the pattern of aggregate tax burdens, studies have attended to overall tax payments from major tax categories as measured by the percentages of total revenue or GDP (Ishi 2001, 8-9;Dalsgaard and Kawagoe 2000, 53;Dewit and Steinmo 2002, 162;Kim 2006, 19, 22;Yang and Min 2013, 60, 74;Kim, M.-K. 2018, 200;Lee 1991, 198-200). This approach, however, produces only indirect tax measures since it calculates the tax figures against broad macroeconomic and fiscal parameters rather than the exact tax base. OECD economists have explored alternative measures by resting on the notion of average effective tax rate (AETR) (Dalsgaard and Kawagoe 2000, 54, 57;Dalsgaard 2000, 35). Taking into account the tax payments evaluated against real income and consumption sources, the scholars have produced more direct measures of tax burdens. This advanced approach, however, is not free of estimation errors. It can produce a wide range of tax figures depending on how the tax base is defined, which may vary across empirical studies owing to their diverse understanding of what constitutes taxable income and consumption (Carey and Rabesona 2003;Carey and Tchilinguirian 2000).
Considering these advantages and disadvantages from existing measures, the present study utilizes both the AETR and GDP-ratio figures to configure a comprehensive pattern of tax burdens across countries. The analysis focuses on four major tax categories: wage income, capital income, SSC, and consumption. Wage income constitutes the wage portion of personal income. Capital income includes corporate income, the non-wage portion of personal income (including mixed income, interests, dividends, and investment income), property, and other incomes accruing from capital transactions. SSC concerns all contributions to public insurance from employers and employees. Consumption covers the spending on general and specific goods and services. Scholars have produced several AETR formulae for these tax categories, covering various years up to 1988 (Mendoza, Razin, and Tesar 1994), 1997 (Carey and Tchilinguirian 2000), 2000 (Carey and Rabesona 2003), and 2015 (Park 2020). Appendix 1, provided online, discusses technical details of seven AETR formulae for wage income and SSC, five formulae for capital income, and three for consumption employed in this study. These formulae are replicated to extend their data up to the year 2018, which are then averaged to produce an overall AETR score for each tax category. The same procedure is repeated by replacing the denominators of the AETRs with the GDP, to determine whether the original pattern of tax burdens identified with the AETRs continue to hold with alternative figures calculated with the broader denominator. All these figures are standardized against respective OECD averages for cross-national comparison.
To gauge the extent that direct taxes contribute to public revenue vis-à-vis non-direct taxes, the author presents a composite indicator that compares all the standardized figures of AETRs and GDP ratios with each other. First, the sum of the standardized AETRs for SSC and consumption (non-direct taxes) are deducted from the sum of wage and capital incomes (direct taxes). The same procedure is repeated for the GDP-ratio figures. Thereafter, these two figures, AETRs and GDP ratios, are averaged to produce the final composite score for the Dependence on Income Taxes. If this score is greater than zero, direct taxes take more burdens to support the general government revenue, and if the score is below zero, non-direct taxes assume more burdens.
Another important dimension of the tax base, exemptions and deductions on taxes, is measured by the differences between formal tax responsibilities and actual tax payments. Close attention is paid to the wage and capital income taxes to which most tax expenditures are directed. Case studies have so far examined detailed tax data to examine the actual amounts or percentages of the tax payments reduced by various exemptions and deductions (Ishi 2001, 94-103, 187-196;Kim, D.-K. 2018, 66-76). These specific figures, however, are difficult to replicate in a cross-national setting owing to the lack of common standards to translate country-level information to internationally comparable statistics.
A practical solution is to compare the standardized statutory rates of wage and capital income with their effective rates. Given the absence of representative statutory rates for wage and capital income taxes, the marginal rates for the highest brackets of personal and corporate incomes are used as proxies. The effective rates are measured with the AETRs and GDP ratios. These statutory and effective figures are then standardized against respective OECD averages, and their differences are calculated by deducting the effective rate from the statutory rate for each of the income taxes. Finally, these differences are averaged to produce a composite index of Income Tax Deduction and Exemption. To the extent that the statutory rates are greater than the effective rates, the index will produce larger positive scores, which indicates that exemptions and deductions are widely available for income tax payers. Table 3 summarizes all tax indicators discussed so far. Appendix 2, provided online, describes their cases, coverage, and data sources.

Empirical evidence
This section summarizes the results of the quantitative measurements for Japan and Korea, in comparison with 19 OECD countries from 1995 to 2018. These years were selected based on the availability of the revenue and national account statistics from the OECD database (https://stats.oecd.org/). Japan and Korea began major tax reforms in earlier years with the introduction of the VAT and the democratic transition in the late 1980s, respectively (Ide and Steinmo 2009;National Assembly Budget Office 2018, 63-65). However, as demonstrated in the following analysis, the second half of the 1990s retained all major features of their previous model because it was still an early phase of major reform. Supplementary tax data as far back as the 1980s lend support to this point (not reported in Tables 4-6, but the data sources are described in Appendix 2). In addition, various government documents and expert assessments corroborate key quantitative findings, particularly those pertaining to more recent developments in Japan and Korea. A high positive score for Indicator 1 means that direct taxes take more significant burdens for the general government revenue than non-direct taxes. A high positive score for Indicator 2 shows that exemptions and deductions are widely available for income taxes.
Before reviewing the empirical results, it is worth mentioning that the two Asian countries present fully comparable cases for public policy analysis with other advanced democracies. Apart from the political democracy that both countries have maintained since the 1950s and the late 1980s, they have achieved successful economic development. The GDP per capita in Japan (PPP-adjusted) was $25,488 during the 1995-2000 period, higher than the average of the 19 OECD countries examined in this study: $24,745 (https://stats.oecd.org/). The Korean figure, $15,703, was lower than this average but nevertheless close to $17,487, the average of the three low-bound countries from Southern Europe (Greece, Portugal, and Spain). In 2011-2018, the Japanese figure was $39,792, moderately lower than the OECD average of $46,912. The Korean figure, $37,074, was also lower, but considerably higher than the average of the three Southern European countries: $30,709. All these features imply that Japan and Korea share broad political and economic contexts with other OECD countries and are thereby bound by similar scope conditions for comparative policy analysis. Table 4 summarizes the general government revenue as a percentage of GDP, covering the 19 OECD countries, plus Japan and Korea. The left column displays the revenue data for the 1995-2000 period. The right column illustrates data for a more recent period, 2011-2018, following more than a decade of tax reforms in the 2000s and 2010s. The 19 OECD countries are classified into several groups with similar tax levels, by applying a standard technique of hierarchical cluster analysis. Close attention is paid to Clusters 1 and 2, which respectively represent the lowest tax group and the one with higher taxes whose level was nonetheless lower than the OECD average. Cluster 1 satisfies a key empirical requirement for residualism, whereas Cluster 2 does for constrained activism.

Total taxation
Ireland presents an interesting case here. Not only did the tax level display a significant change between the two periods, but the tax model itself exhibited unique elements that did not fit with other low-tax cases. In the past, Ireland mimicked the UK model of constrained activism by featuring regressive taxation combined with a limited room for exemptions and deductions. However, unlike the UK case, public revenue relied more heavily on consumption taxes, resulting in non-direct taxes accounting for a major share than direct taxes in public revenue (O'Toole and Cahill 2006, 206-207;Sommacal 2004, 161-165, 173-174). Since the 2000s, total taxation has been cut down to a level comparable to the residualism cases. Similar to these cases, exemptions and deductions were tightly restricted. However, taxation continued to be regressive. Non-direct taxes assumed even larger burdens than before, owing to aggressive cuts in corporate taxes (O'Toole and Cahill 2006, 208;Sommacal 2004, 176-177). Confirmed with various tax data in this section, these features made Ireland a unique case for low-tax countries.
Next to Japan and Korea, these countries presented empirical results consistent with the notion of developmentalism during 1995-2000. Japan displayed a low level of 25.8 percent with a standardized score being -1.6, which was marginally below the average of Cluster 1 countries. Korea exhibited a much lower figure, at 19.3 percent, with a standardized score of -2.5. Upon entering the 2011-2018 period, however, total taxation in Japan approached Cluster 2 cases. The Korean figure also increased considerably to close the gap with the residualism cases.  Table 5 presents several statutory and marginal effective tax rates to assess the progressivity of taxation. It covers Japan and Korea along with Cluster 1 and 2 countries. It first summarizes the top statutory rates for personal and corporate income taxes. These figures are standardized against respective OECD averages and then averaged to produce a composite statutory index, The Average of Top Statutory Rates. In the last column, the marginal effective rates for the families with 167 percent of the average gross wage earnings are reported as a proxy for the real tax responsibility for highincome earners. Both these statutory and effective measures are compared with the standardized VAT rate to gauge the degree of tax progressivity. Residualism countries in Cluster 1 showcase progressive taxation. In five of six cases (two from the 1995-2000 period and four from the 2011-2018 period), both the top statutory and marginal effective rates were higher than the VAT rates. The only exception was Ireland in 2011-2018, which produced an indeterminate outcome. 2 Notice that the USA did not have a federal-level VAT policy. Its VAT data therefore were calculated by averaging out all state-level VAT rates. Meanwhile, Cluster 2 countries produced diverse results. In Australia (1995Australia ( -2000 and Canada (2011Canada ( -2018, like in other residualism cases, the statutory and marginal effective income tax rates were all higher than the VAT rate. However, regressive taxation was dominant in the UK, Ireland (1995, and Southern European countries. Seven of the eight countries in this group displayed regressive outcomes, while Portugal (2011Portugal ( -2018 showed only an indeterminate case.

Progressivity of taxation
Next to the two East Asian countries, Japan presented progressive taxation that was consistent with fiscal developmentalism. While there was a notable difference between the statutory and effective marginal income tax rates as a result of extensive exemptions and deductions for income earners, both tax figures exhibited higher standardized scores than the VAT rate during the 1995-2000 period. Back in the 1980s, while no standardized data were available, the statutory rates for personal and corporate income were even higher at 66.4 percent and 54.2 percent, than the 3 percent VAT rate introduced in 1989. A similar pattern continued for the 2011-2018 period, despite a gradual increase in VAT up to 10 percent in 2019 (Ishi 2001, 324-339;Nippon Times 2019).
Expert studies support this assessment by emphasizing the role of progressive corporate taxation in Japan, whose rates have been at all times higher than the OECD average, in contrast to low consumption tax rates (Brownlee and Ide 2017;De Mooij and Saito 2014, 6-8;Ide and Steinmo 2009, 122-128;Keen 2008, 64). The Cabinet Office and public officials have also shared this assessment since the late 1980s (Ishi 2001, 324-341;Tax Commission 2003, 9-10, 2007Cabinet Office 2014, 30, 2011a, 5, 2011b, when they started to cut down corporate taxes whose high rates were deemed a drag on the Japanese economy. The Korean story is more complicated. The statutory income tax rate was higher than the VAT for both periods. The difference was even greater in the 1980s, with the personal and corporate rates set at 61.1 percent and 38.1 percent, respectively. However, the standardized marginal effective rate, set at a rock-bottom level among all OECD countries including Japan, was consistently lower than the standardized VAT. Despite these indeterminate results, Korea still presented a unique case of progressive taxation. Such an assessment had much to do with the extraordinarily progressive nature of income taxation in Korea. As much as higher income earners enjoyed wide deductions and exemptions, medium-to-low income earners were awarded even wider tax benefits. For instance, during the 1995-2000 period, families with 67 percent of the average gross wage earnings were subject to only 18.3 percent of the marginal effective income tax rate set for the 167 percent income group (self-calculation using the OECD data from https://stats.oecd.org/). This was the lowest figure, considering that the average score for Cluster 1 and 2 countries was as high as 61.2 percent (close to the OECD 19 average of 61.3 percent). In the 2011-2018 period, although the Korean figure improved to 40.2 percent, it was nevertheless at the lowest level, with the average score for Clusters 1 and 2 being 62.9 percent (close to the OECD 19 average of 62.8 percent). A vast majority of income earners were largely relieved of tax responsibilities as they paid only low consumption taxes.
This point is echoed by expert studies that highlight various deductions targeting medium-to-low wage earners, as well as wide exemptions that have allowed about a half of total income earners to pay zero taxes (Ahn and Oh 2018, 22;Jeong 2018, 215;National Assembly Budget Office 2017, 108-110). The authoritarian government until the late 1980s had instituted the foundation of these tax benefits (Ahn 1997, 280;M.-K. Kim 2018, 177-178;Kim 1991, 444-445). Democratic governments have then built on the same idea to introduce various new benefits. Examples include the deductions and exemptions targeting public pension contributions, medical expenses, education expenses, and credit card payments (MOEF 2001(MOEF , 49, 69, 71, 2002(MOEF , 25, 27, 47-48, 2003; home loans, income for low-paid workers, and supplementary retirement pension contributions (MOEF 2004(MOEF , 77, 78, 80, 86, 93, 2005(MOEF , 42, 44, 2007; and maternity payments, child allowances, earned income tax credits, monthly rental payments, wages and employment in small to middle-size companies, and single moms (MOEF 2007(MOEF , 273-283, 2008(MOEF , 105-106, 287-290, 2009(MOEF , 65, 68, 326, 331-333, 2010(MOEF , 98, 112, 2011(MOEF , 106, 141, 2012(MOEF , 28, 35, 53-56, 2013. Table 6 presents two measures of aggregate tax burdens, AETR and the GDP ratio, calculated for wage income, SSC, capital income, and consumption. Based on the configuration of these figures, a composite index called the Dependence on Income Taxes was produced to quantify the degree to which public revenue relies on direct income taxes vis-à-vis non-direct taxes. The last column reports another composite index, Income Tax Deduction and Exemption, which operationalizes the pervasiveness of tax expenditures for wage and capital incomes. Countries in Cluster 1 present a sound case for fiscal residualism. In five of the six cases, direct income taxes assumed higher burdens than non-direct taxes. Ireland in 2011-2018 was an outlier where non-direct taxes had higher burdens. Meanwhile, all negative indications for the Income Tax Deduction and Exemption index, except the US in 2011-2018, suggest that the scope for tax expenditures was relatively restricted for income tax payers from the OECD standpoint. 3   [1995][1996][1997][1998][1999][2000] and Canada (2011Canada ( -2108, while exhibiting similar features with Cluster 1 countries, tended to maintain tighter restriction on tax expenditures, as shown by the lower scores in the Income Tax Deduction and Exemption index. Direct taxes, demonstrated in the Dependence on Income Taxes index, also tended to take heavier burdens than was true for most Cluster 1 cases. The UK shows another similar case that combined broad direct taxes with a tight control of tax expenditures. Meanwhile, Southern European countries were a different case where non-direct taxes (SSC and consumption taxes) assumed a greater role in public revenue than direct income taxes. Taxpayers taking on those tax burdens were instrumental for compensating for extensive exemptions and deductions embedded in income taxes. Lastly, Ireland (1995 mimicked the UK by limiting the room for exemptions and deductions but, unlike the UK, relied more heavily on non-direct taxes, particularly, consumption taxes. All these results are consistent with the theoretical discussions provided in the previous sections.

The tax base
Meanwhile, Japan and Korea showcase interesting stories that cut across diverse tax models. During the 1995-2000 period, Japan provided a sound example for fiscal developmentalism. A high score in the Income Tax Deduction and Exemption index suggested that the income tax base was maintained very narrow. Despite this constraint, direct income taxes (driven by capital taxes) took higher burdens than nondirect taxes. This was true even considering the sizable SSC that supported general health and pension programs that had been steadily expanded during the post-war period (Campbell 1992;Estévez-Abe 2008). Historical data from the 1980s support this assessment. While standardized statistics were not available due to the lack of comparable OECD data, a similar pattern of tax burden was observed. Capital income assumed the highest burden (38.6% for the AETR and 10.1% for the GDP ratio), followed by SSC (13.2% and 7.5%, respectively). Meanwhile, wage income (10% and 5.7%, respectively) and consumption (6.1% and 3.5%, respectively) were lagging further behind.
Japan introduced notable changes in its tax base during 2011-2018. The income tax base continued to be narrowly defined, as shown by the persistent large score in the Income Tax Exemption and Deduction index. However, non-direct taxes began assuming more burdens than direct taxes. The score of the Dependence on Income Taxes index declined significantly from 1.2 to 0.0. If we look more closely at 2015-2018, the score was even negative at -0.2 (not reported in Table 6). Obviously, this result reflects important changes in the tax policy in Japan where, as reported in Table 6, SSC expanded to become the largest contributor to public revenue, even surpassing capital taxes whose contribution consistently declined. Consumption taxes added more weight to non-direct taxes, although they continued to be the lowest contributor to public revenue. Overall, Japan presents a new case of constrained activism that features more contributions from non-direct taxes-mostly by SSC-along with a large scope for exemptions and deductions for income taxes.
Historical and public documents support these assessments, demonstrating that 50-60 percent of wage income was not subject to taxes in various years during the 2000s and 2010s (Jones and Tsutsumi 2008, 23;OECD 2018, 11). This can be attributed to extensive tax benefits available for low income earners, home-related loans, dependent families, and others (MOF 2013(MOF , 1, 2018Tax Commission 2007, 13). Corporate income taxes also continued to be largely exempted or deducted with special treatments on R&D, fixed capital investment, corporate restructuring, wage supports, family businesses, local companies, and small and medium-sized companies (Jones and Tsutsumi 2008, 17, 19;MOF 2013MOF , 3, 2016MOF , 3-4, 2017MOF , 3, 2018. Thereafter, a broad neoliberal consensus among politicians toward "innovative growth" (Cabinet Office 2007, 2), employment-friendly tax cuts (Cabinet Office 2011a, 5), international competitiveness (Cabinet office 2014, 30), and "economic revitalization" (Cabinet Office 2015, 29) led the Japanese government to further reduce the effective tax rates for corporate income (MOF 2010(MOF , 3-4, 2015a(MOF , 2, 2015b. Under these circumstances, government officials sought alternative revenue sources to support their expanded spending. Various studies and public reports highlighted a sharp increase in SSC in the post-2000 period (Chopel, Kuno, and Steinmo 2005, 28-29;Kodama and Yokoyama 2018, 993;Tax Commission 2007, 4). The VAT rate also increased in a phased manner in order to compensate for revenue losses driven by capital tax cuts (Brownlee and Ide 2017, 68-74;De Mooij and Saito 2014, 30;Ide and Steinmo 2009).
Korea presents another dynamic story for the tax policy. The country featured a strong case of developmentalism during the 1995-2000 period. The score in the Income Tax Deduction and Exemption index was greater than those reported for the residualism cases. Subsequent losses in public revenue were partially compensated for by non-direct taxes and particularly by consumption taxes that assumed the highest burden of all major tax categories. Again, the level of consumption tax burdens was not high from the OECD perspective-a point validated by the low standardized AETR and GDP-ratio scores in Table 6. However, wage taxes, capital taxes, and SSC all took on much lower burdens, even less than all residualism cases, owing to the near absence of public services/assistance and general insurance programs (Yang 2017). Historical data from the 1980s lend additional support to this assessment. Although standardized figures were not available, consumption was the largest contributor to public revenue with 13.5 percent of the AETR and 8 percent of the GDP ratio, followed by capital income (with 11.0 percent and 4.0 percent), wage income (with 3.3 percent and 1.7 percent), and SSC (with 1.0 percent and 0.5 percent).
Over the last few decades, Korea experienced notable changes in its tax base to approach the residualist model. First, exemptions and deductions were significantly reduced. The Income Tax Deduction and Exemption index, whose score used to be 0.6 in the 1995-2000 period, turned relatively negative at -0.01. Although this score was nevertheless higher than what was observed for most of the residualism cases, the gap was substantially closed. Aggregate tax burdens also exhibited significant changes. While wage income continued to bear the smallest burdens, capital income became the largest contributor to public revenue, thereby surpassing the standardized scores of consumption taxes and even the SSC, whose aggregate contribution expanded significantly toward a level comparable to the residualism cases. This observation is clearly supported by the GDP-ratio statistics in Table 6. The AETR figures, though, exhibit a rather vague pattern with the standardized scores of SSC, capital income, and consumption being -0.8, -1.0, and -1.0, respectively. However, such uncertainty is significantly reduced in the second half of the 2010s.
During 2015-2018, those scores were -0.74, -0.79, and -1.0, respectively. The GDP ratios presented an even more favorable configuration for capital taxes, with the standardized scores being -0.7, 0.1, and -1.2, respectively (not reported in Table 6). While AETRs became virtually indistinguishable between SSC and capital income, the GDP ratios clearly supported a leading role by capital income. Reflective of these configurations, overall direct taxes began assuming a more prominent role in public revenue. The Dependence on Income Taxes index, whose score was -0.6 previously, reached -0.1 during 2011-2018. A closer analysis of the 2015-2018 period showed that the score was even positive at 0.2.
Expert assessments and public documents support these estimates. About a half of income earners in Korea continued to pay zero taxes in the 2010s (Ahn and Oh 2018, 22), although the government made efforts to expand the tax base for high-income earners by reducing various deductions and exemptions (Jeong 2018, 215) and also identifying further income sources for profitable family and self-employed businesses (Jeon and Jeong 2012;Lee 2012, 514). Meanwhile, the capital income tax base was greatly expanded by reducing tax expenditures for businesses (in large parts, established businesses) and also tightening legal control over the procedure with and the total amount of such expenditures (Kim, H.-S 2012, 247, 249;MOEF 2014MOEF , 285, 288, 293, 305, 2015MOEF , 63-64, 136, 152, 2017. The tax base for other financial and property incomes was also expanded. In the early 1990s, sweeping reforms in the financial and property markets provided the Korean government with an efficient tax-collecting power, particularly by requiring all citizens to use their real names in financial and property transactions (Ahn 1997, 283-305). Since then, left-leaning governments have expanded the capital tax base more actively than their right-leaning counterparts, and especially those without budgetary discipline (Park 2022;Kim, H.-A 2012, 111;Lee 2012, 513, 551-552).
Reflective of these changes, aggregate tax burdens for capital income notably increased, particularly in contrast to wage income and consumption, whose trends were rather stagnating (Ahn 2012, 47-48;Lee 2012, 509). SSC also increased considerably (Ahn 2012, 48;Lee 2012, 511), following the introduction of general pension and health programs in the late 1980s. Successive welfare reforms that aimed at slim but broader benefits (Yang 2017) further encouraged the Korean government to expand the SSC by facilitating rate hikes or incorporating new contributor groups into the financing system (National Assembly Budget Office 2019, 17, 53, 85-88). While building on these assessments, the present study demonstrates that capital income taxes and thereby direct income taxes achieved more impressive expansion, eventually outpacing SSC and non-direct taxes, which also registered significant expansion.

Conclusion and discussion
This study explored key features of the old and new tax models in Japan and Korea in comparison with two alternative models, residualism and constrained activism, which the author drew from low-tax OECD countries. It reconfirmed that Japan and Korea undoubtedly presented a strong case of developmentalism along with lowest but progressive taxation and a wide range of exemptions and deductions for income tax payers. Within this framework, the two countries developed diverse features in the level of total taxation and the pattern of tax burdens between direct and non-direct taxes. In recent years, Japan and Korea have become more comparable to other OECD cases. Japan now belongs to a group of constrained activism. Korea is making a transition to the residualism model. All these assessments were validated by extensive tax data and other supplementary documents that address various dimensions of the tax structure from the 1980s to 2018.
Conducting a thorough empirical and conceptual analysis of the tax models in Japan and Korea, this study sheds light on relevant but under-explored areas for further research. First, it opens a discussion for a policy-level causal mechanism that may exist in tax adjustments in Japan and Korea. Accordingly, it is interesting to find that a rise in SSC-which was driven by such inevitable forces as the maturation of public insurance programs and population aging (Dalsgaard and Kawagoe 2000, 10;Chopel, Kuno, and Steinmo 2005, 28-29;Yang 2017)-seems to have limited the scope for further expansion in wage income taxes. Capital taxes, which have been known to be influenced by various forces such as government partisanship, political and economic institutions, and the legacies of existing tax policy (Hallerberg and Basinger 1998;Cusack and Beramendi 2006;Ganghof 2007;Hay 2003;Shin 2017;Swank 2006Swank , 2016, may also have left a considerable impact on other tax categories. Considering the difficulty with increasing wage income taxes at a time of rising SSC, their impacts may have been channeled through consumption taxes. Effectively, cuts in capital taxes in Japan were accompanied by a gradual increase in consumption taxes. In Korea, the expansion of capital taxes coexisted with the stagnation of consumption taxes. It will be interesting to explicate such balancing mechanisms with fine-tuned causal accounts, whereby changes in certain tax categories facilitate adjustments in others. This will certainly expand our knowledge regarding how institutional complementarity or hierarchy works at the policy level. It will also contribute to existing studies on Japan, Korea, and other OECD countries where researchers have mostly focused on macro-level mechanisms of policy adjustment whereby the tax policy interacts with broad economic structures (coordinated market economy vs. liberal market economy) and public welfare regimes (Beramendi and Rueda 2007;Cusack and Beramendi 2006;Kato 2003;Kemmerling 2009, 17;Kim 2013;Swank 2016).
Overall, this study promotes an integrative analysis of comparative public policy in East Asia and beyond. Considering that tax policies are closely intertwined with the spending policies they support, recent changes in tax policies of Japan and Korea indicate that these countries are experiencing a broad transformation in their public policy regimes. Studies have already explored the spending side of the change by tracing new features of public welfare policy, such as more inclusive public insurances and newly expanded family, care, and gender-egalitarian programs, which have grown to a comparable level with other OECD cases (Choi, Fleckenstein, and Lee 2021;Yang 2017). The present study adds novel evidence to the revenue side of the change. Accounting for all these developments will require more comprehensive causal equations. It will also be interesting to examine what are responsible for these increasing similarities with those for general OECD cases and what explains the still remaining differences.