MPIfG Working Paper 06/1, January 2006
The Politics of Tax Structure
, Max Planck Institute for the Study of Societies Cologne
An earlier version of this paper was presented at the Yale Conference on
Distributive Politics, April 29-30, 2005. The paper is now under review for an
edited volume to be published by Cambridge University Press. I would like to
thank the conference participants, especially Kathleen McNamara and Michael
Wallerstein, as well as the in-house referees, Andreas Broscheid and Armin
Schäfer, for helpful comments and discussions. All remaining errors are mine.
Governments that wish to redistribute through budgetary
policy do so mostly on the spending side, not on the taxation side of the budget.
The taxation side is nevertheless important, partly because less efficient tax
structures seem to be associated with lower taxation and spending
levels. Hence political conflicts over spending levels may partly be fought
as conflicts over tax structure. The paper provides a coherent perspective on
the politics of tax structure. Specific topics include the (ir-)relevance of tax
mixes, policy change in income taxation, the importance of tax competition, and
the role of political institutions.
Wenn der staatliche Haushalt zur Umverteilung genutzt
wird, so geschieht dies vor allem auf der Ausgabenseite und nicht auf der
Einnahmenseite. Die Einnahmeseite ist trotzdem bedeutsam, weil es offenbar einen
Zusammenhang gibt zwischen der Effizienz der Steuerstruktur und dem
Steuerniveau. Daraus folgt, dass politische Auseinandersetzungen über
Ausgabenniveaus teilweise als Auseinandersetzungen über die Steuerstruktur
geführt werden können. Das Papier entwickelt eine kohärente Perspektive zur
Analyse von Steuerstrukturpolitik. Zu den behandelten Themen gehört die (Ir-)Relevanz
des Steuermix, der Politikwandel in der Einkommensbesteuerung, die Bedeutung des
Steuerwettbewerbs und die Rolle politischer Institutionen.
Efficient tax structure and differentiated income taxation
Tradeoffs and trends
The "spill-over problem" and the role of tax competition
The domestic politics of differentiated income taxation
Governments that wish to redistribute through budgetary policy do so mostly on
the spending side, not on the taxation side of the budget. The taxation side is
nevertheless important, partly because a less efficient tax structure seems to
be associated with lower taxation and spending levels. Hence political conflicts
over spending levels may partly be fought as conflicts over tax structure (for
example, Przeworski 1999: 43). A recent example of this logic is the Wall
Street Journal editorial (November 20, 2002) that complained about the low
income tax burden of a US taxpayer earning $12,000. These "lucky duckies" benefit so much from the progressive income tax structure that they pay little
in income tax: "It ain't peanuts, but not enough to get his or her blood boiling
with tax rage." If much less progressive income taxation could somehow be
entrenched, so the argument goes, this would increase voters' overall resistance
to taxation and thus reduce tax levels.
While in this example tax structure "efficiency" refers to the electoral costs
of taxation (cf. Hettich/Winer 1999), the general logic extends to economic and
administrative costs. In continental European welfare states, for instance,
which suffer from a high tax burden on low-skilled labor, progressive taxation
of wages may be a matter of economic efficiency (Scharpf 2001). Similarly, the
vertical distribution of the tax burden is only one dimension of tax
equity and efficiency. The horizontal dimension has to do with the
relative tax burdens on different types rather than levels of income. Along this
dimension it is widely believed that moderate tax burdens on (certain types of)
capital are efficient, both in general and within the income tax system (for
example, De Long/Summers 1991; Przeworski/Wallerstein 1988; Lindert 2004).
For example, the "social democratic" welfare states of Finland, Norway, and
Sweden operate so-called dual income taxes (DITs) which discriminate against
wages in very systematic and visible ways: while capital income is taxed at a
uniform low and proportional rate of 25-30% (or even lower), wages are taxed
progressively up to top rates of almost 60%. Many observers find this difficult
to square with those countries' traditional redistributive ambitions. Former
Conservative Prime Minister of Norway Kaare Willoch (1995: 179), for instance,
laments that while "social democrats once felt that financial income should be
taxed more severely than income from work, they have now changed their minds";
and he is puzzled that the new tax policy goes together with continuing "demands
for a large public sector." But these two observations may be causally linked
rather than contradictory. For if moderate capital taxation increases the
overall (that is, electoral, economic, and administrative) efficiency of tax
structure, it also tends to increase the level of public spending.
In this paper I explore the politics of tax structure in comparative
perspective, drawing on both quantitative and historical evidence. My focus is
mainly on advanced EU and OECD countries and on the potential tradeoffs
between vertical and horizontal tax structure efficiency. These tradeoffs
seem most salient in income tax policy: If horizontal efficiency does indeed
require moderate and proportional taxation of (some) capital income, whereas
vertical efficiency (and equity) considerations require progressive taxation of
wages, the resulting tax rate differentials will give high-income taxpayers
great incentives to transform highly taxed income into lowly taxed income.
Trying to avoid this is difficult and costly in administrative terms, which in
turn tends to decrease political support for tax rate differentiation. Finding a
balance between horizontal and vertical tax structure efficiency is thus
inherently difficult. I will explore the tradeoffs involved and show that this
exploration can inform and connect a number of separate literatures on the
political economy of taxation. Moreover, I will analyze how corporate tax
competition has exacerbated domestic tradeoffs and how political institutions
figure in the domestic politics of income tax structure.
The next section clarifies some basic concepts. Section 3 treats differentiated
income taxation as an implication of the more general need to moderate the tax
burden on capital income. Section 4 highlights basic tradeoffs in income
taxation and interprets the tax reforms of the 1980s and 1990s as efforts to
find systematic and efficient forms of differentiated income taxation. Section 5
shows how tax competition has exacerbated the economic and administrative
problems of differentiated income taxation. Section 6 explores the politics of
differentiated income taxation and the role of political institutions.
In the political science literature the terms "income tax" and "consumption
tax" are typically understood in an institutional sense, following the usage
in, say, the OECD Revenue Statistics: An income tax is a direct tax on
corporations and individuals, while consumption taxes are indirect taxes (for
example, value-added taxes). However, there is also an analytical meaning
of these terms, which is unrelated to the method of implementation. A
consumption or expenditure tax in this analytical sense can be implemented as a
direct and progressive tax on corporations and individuals.
So what is the analytical difference between income and expenditure taxes?
Answering this question requires an analytical distinction between two types of
what we typically call "capital income" (Slemrod/Bakija 2004: 203-204). The
normal return to capital is the return to deferring consumption, the (risk-free)
return to wealth. If I buy a machine, I want this investment to generate at
least the return I would have received from buying government bonds. The normal
return is the return that can be earned on a marginal investment in capital,
which is competed down to a fairly low level. Above-normal returns go beyond
that level. They include various things, for example, returns to innovation,
returns to establishing a monopoly in some market, or returns to entrepreneurial
skill. Bill Gates's income from his share of Microsoft profits would typically
be labeled "capital income," but most of it consists of above-normal returns.
Based on this distinction, we can characterize, roughly, three analytical types
of taxes: A wage tax taxes neither normal nor above-normal returns; an
expenditure tax exempts the former but taxes the latter; and an income
tax taxes both. All three, of course, also tax wages.
These distinctions are important because economic theory suggests that - at
least in closed economies - above-normal returns can be taxed without distorting
investment decisions, whereas taxation of normal returns discourages saving and
investment (for example, Przeworski/Wallerstein 1988). Many also believe that
there is an equity rationale for taxing normal returns at a concessionary rate
or not at all. Since the normal return is a reward for the saver's
willingness to postpone consumption, taxing it (highly) would result in
over-taxation of individuals with a relatively high savings propensity. In
contrast, above-normal returns often represent windfall gains which ought
to bear a high tax rate. Proponents of direct expenditure taxes (that is,
exemption of normal returns) also emphasize that these can achieve exactly the
same overall degree of progressivity as income taxes.
Efficient tax structure and differentiated income taxation
There is widespread agreement that tax structures and tax levels are
interrelated; more efficient tax structures are associated with higher tax
levels (Hettich/Winer 1999). "Structure" is here meant to include both the mix
of different types of taxes and the internal structure of particular taxes. "Efficiency" is understood broadly, referring to the economic, political (electoral), and
administrative costs of taxation. Some authors highlight the causal arrow from
levels to structures. The argument is that, at least in democracies, a high
taxation level enforces an efficient tax structure (Lindert 2004). Others
highlight the opposite causal arrow: changes in tax efficiency entail changes in
taxation and spending levels (Becker/Mulligan 2003; Kato 2003).
There is less agreement about what makes for an efficient tax structure. For
instance, some see flat tax rates as crucial to tax structure efficiency (for
example, Becker/Mulligan 2003), while others make efficiency-based arguments in
favor of (directly) progressive taxation (for example, Frank 1999; Layard 2005;
Røed/Strøm 2002). If there is an exception to these kinds of disagreements, it
probably concerns the normal return to capital: Since the economic, political,
and administrative costs of taxing capital income seem to increase quickly,
large welfare states may be predicated on moderate capital income taxation.
Following Przeworski and Wallerstein (1982, 1988), many political scientists
adopted this perspective. It was argued that left-wing governments generally
kept the tax burden on (normal) capital income low in order to reconcile
efficiency on the taxation side of the budget with equity on the spending side (for
example, Garrett/Lange 1991). In economics, Lindert (2004) advanced a similar
argument as one explanation of the relatively good economic performance of large
tax/welfare states. While these arguments focus on the average effective
tax burden on capital, a focus on marginal rates leads to parallel
results: taxing (normal) capital income with graduated and high marginal tax
rates seems inherently problematic (Slemrod 1990).
[click on table to enlarge view]
These arguments imply that the ratio of capital taxes to labor and consumption
taxes decreases systematically with the total tax burden. To show that this is
the case in advanced OECD countries, I combine data from Eurostat (2004) and
Carey/Rabesona (2002) for the period 1995 to 2002 (Table 1).
Figure 1 shows the correlation between the total tax burden (revenue as % of GDP)
and the capital/(labor+consumption) ratio after controlling for the degree of
wage coordination. This control is included because it has been argued that in
countries with corporatist decision-making, the efficiency costs of labor income
taxes are reduced so that labor is taxed more heavily (Summers/Gruber/Vergara
1993; cf. Cusack/Beramendi 2003).
As expected, the ratio of capital taxes to labor and consumption taxes clearly
decreases with total taxation.
The impact of
total tax levels on the relative importance of capital taxation
Notes: Unweighted averages for the period 1995–2002.
Sources: See text and Table 1.
How do we get from the assumption that tax efficiency requires a moderate
overall level of capital taxation to the observation of differentiated forms of
income taxation? Some political scientists seem to think we do not get there at
all because they understand tax structure efficiency solely in terms of
tax mixes. Their argument, as I understand it, is this: Since
"progressive" property and income taxes fall heavily on capital, whereas
"regressive" consumption and social security taxes do not, a high share of the
latter in a country's tax mix is conducive to building and maintaining large
tax/welfare states (Wilensky 2002: 392; Kato 2003: 199).
I think this argument is flawed precisely because it ignores the variability of
the internal structure of income tax. Low income taxation is only one of two
ways of keeping the capital income burden moderate; the other is differentiated
income taxation which systematically discriminates against wages within
the income tax system.
Consider the comparison between France and Denmark (all values are unweighted
averages for the period 1995-2002 - see Table 1): France seems to have a very
"regressive" tax mix (Kato 2003: 94-110), with a corporate and personal income
tax burden of only 9.8% of GDP. In contrast, Denmark has a huge income tax
burden of 28.8% of GDP. Based on the argument from tax mixes, the large size of
Denmark's tax/welfare state is a major puzzle (Kato 2003: 197). Yet Denmark's
outlier status disappears once we focus on the underlying tax burdens on capital
and labor (Ganghof 2007 forthcoming). Despite its low income tax burden, France
has an average tax rate on capital (income and stocks) of 35.5%, the highest
in the EU, whereas Denmark has one of only 30.5%. The reason for this is
Denmark's highly differentiated income tax: Employed labor is taxed at an
average effective rate of 41%, with a top marginal rate that exceeds 60% and
sets in below the earnings level of an average production worker (Table
1). In contrast, many important types of capital income are taxed at marginal
rates of 30% or lower, which contributes to an average effective tax burden on
capital income of only 20.1% (France: 18.8%).
Hence the difference between Denmark and France does not lie in the level of
capital taxation but rather in the progressivity of wage taxation. In France the
dominant form of direct wage taxation is payroll taxes, while in Denmark it is
income taxes. Payroll taxes, regarded as taxes, are typically regressive because
they do not have basic tax allowances, which reduce the relative tax burden on
low-income earners, but do have ceilings on contributions, which reduce the
relative tax burden on high-income earners. Therefore, countries with "small" income taxes such as France or Austria have also adopted differentiated income
taxes that apply fairly high marginal tax rates to wages while treating
sensitive types of capital income under separate and much lower rates. In these
countries, the income tax partly fulfills a kind of progressivity adjustment
function for wage taxation at large. Since a large share of labor taxation
takes the form of payroll taxes and consumption taxes, income taxes with
graduated rates and a large basic allowance can provide a progressive
This argument can also help to explain the fact, highlighted in Ganghof (2005b,
2006 forthcoming), that the total tax burden (as % of GDP) is a much
better predictor of top personal income tax rates than the income tax
burden (as % of GDP). The logic is as follows: The difference between high-tax
and low-tax countries is mainly accounted for by direct or indirect taxes on
labor incomes (cf. Figure 1) and a high labor tax burden implies higher top
personal income tax rates on labor incomes - either directly (if the share of
income taxes in total taxes is large) or indirectly (if income taxes are
relatively small but more progressive).
Tradeoffs and trends
In this section I use the assumption of fairly tight constraints on the taxation
of (normal) capital income to offer a stylized interpretation of the post-war
development of income taxes in advanced OECD countries. I argue that policy
change can be understood as the result of efforts to move towards more efficient
forms of tax rate differentiation. This will allow us to better understand the
nature of existing income tax regimes, such as the Nordic dual income taxes, as
well as the effects of tax competition (to be discussed in the next section).
Because my interpretation differs somewhat from the "standard" view in political
science, I summarize this view first.
Political science explanations of income tax reforms in the 1980s and 1990s
highlight the importance of ideational changes. Swank and Steinmo (2002: 645)
speak of a "paradigm shift," Kato (2003: 14) of a "reversal of the ideal of
progressive income taxation." These accounts focus mainly on the structure
of capital income taxation. Privileges for business investment are interpreted
as efforts of "micro-management" on the supply-side (Steinmo 2003) and tax
privileges for personal savings (pensions, owner-occupied housing) as "hidden" welfare provision (Howard 1997; Ervik 2000). Policymakers, so the argument goes,
learned that these manipulations of the structure of capital taxation were
inefficient, so that the tax reforms of the 1980s and 1990s leveled the playing
field and made income taxation more "market-conforming" (Swank 1998; Garrett
While there is much truth to this view, it underestimates the extent to which
tax privileges for capital income were due to the aim of reducing the level
of capital income taxation and the fundamental administrative problems of taxing
many types of capital income. As a result, it also underestimates the extent to
which the very same basic tradeoffs continue to structure the politics of
taxation. To elaborate on these points, I start by reviewing, in a stylized way,
what I consider to be the basic tradeoff faced by policymakers.
As already explained, there are two ideal-type "income" taxes: truly uniform
income taxes, which tax all types of incomes jointly and equally, and direct
expenditure taxes, which exempt the normal return on capital. On the
assumption that taxing all normal returns on capital is very costly, politically
acceptable versions of both ideal-types are likely to have very different tax
rate schedules. If policymakers followed the income tax ideal closely, tax rates
would have to be moderate. Marginal income tax rates would be leveled down to
the lowest common denominator - defined by what is acceptable for the most
"sensitive" (that is, costly-to-tax) types of capital income. In fact, given the
inherent administrative problems of taxing certain types of capital income with
graduated rates, a truly uniform income tax would most likely have a flat tax
rate, with or without a sizeable basic tax allowance (Slemrod 1990; Slemrod
1997). In contrast, if policymakers followed the expenditure tax model, a
progressive rate structure would perhaps imply lower economic efficiency costs
because normal returns are exempt. In addition, more progressive taxation of
above-normal returns and wages would be necessary to make up for this exemption
in equity and revenue terms. Given real-world constraints on capital income
taxation, therefore, policymakers who want to implement a "market-conforming" income tax that does not intervene in the structure of savings and investment
face the basic choice between a flat income tax and a progressive
This choice is a difficult one because both ideal-types maximize certain
dimensions or notions of tax efficiency and equity at the expense of others
(Ganghof 2006, forthcoming). It is not surprising, therefore, that policymakers
in OECD countries eschewed a clear decision and instead tried to combine
elements of both types of taxes. The problem was that the resulting "hybrids" or
"compromises" (Aaron/Galper/Pechman 1988) were ad hoc and proved to be very
inefficient. Two examples: One type of direct expenditure tax at the business
level, the so-called cash-flow tax, allows investment outlays of businesses to
be deducted immediately, thereby exempting the normal return on capital. However,
if this "expensing" of investment outlays - or less extremely: greatly
accelerated depreciation of assets - is combined with the income tax feature of
interest deductibility, as done in the past by some OECD countries, the tax rate
on the normal return to debt-financed investment becomes negative. This
leads to significant distortions in the structure of investments.
Something similar happened with respect to owner-occupied housing, especially in
the Nordic countries: If mortgage interest payments are deductible, a feature of
a pure income tax, but the resulting implicit rent earned by owner-occupiers is
taxed at a low level or not at all, an expenditure tax feature such as housing
investment is highly subsidized (Sørensen 1998). This unsystematic combination
of features of income and expenditure taxes compromised the efficiency, equity,
and revenue goals of income taxes. For instance, it was estimated that in 1986
Danish taxation of personal (household) capital income resulted in a revenue
loss of 1.6% GDP, which was roughly balanced by the revenue from corporate
taxation (Ganghof 2007 forthcoming). Denmark's huge income tax was thus largely
a tax on wages and transfers.
While the tax reforms of the 1980s and 1990s responded, in part, to these kinds
of pathologies, the underlying dilemma did not simply disappear: If policymakers
were unwilling to implement pure income or expenditure taxes, they had to look
for a more systematic and practicable hybrid approach. I argue that this is what
most advanced OECD countries did. One of the best results of this search is the
dual income tax (DIT) model implemented in Finland, Norway, and Sweden, as well
as, temporarily, in Denmark and Italy.
I shall begin by briefly characterizing this model (Sørensen 1998; Cnossen
2000), as this will facilitate the subsequent discussion of the effects of
corporate tax competition.
The DIT model combines features of flat income taxes and progressive expenditure
taxes in straightforward ways (Ganghof 2006 forthcoming: chapters 3 and 4).
Capital income is generally taxed at moderate uniform proportional tax rates
between 25% and 30%, while wages are taxed with graduated rates up to top rates
of around 60%. The above-normal returns of businesses (as approximated by tax
administrations) are, in part, taxed together with wages - as they would be
under an expenditure tax. The reason is this: In order to split business income
into its capital and labor components, tax administrations typically impute a
"normal return" on the invested capital and tax all residual income as wages.
In the Nordic countries, these splitting regimes are only applied to
unincorporated businesses and small corporations in which (some of) the owners
are active as managers. However, the Italian version of the DIT, introduced in
1998, applied the splitting regime to all businesses (but only to the
increase in their net equity after the base year) (Bordignon/Giannini/Panteghini
2001). Firms' normal returns were taxed at 19%, their residual profits either at
the corporate rate of 37% or at personal income tax rates of up to 46%.
After a transitional period, therefore, an ideal DIT of the Italian type would
be a systematic compromise between a flat income tax and a progressive
expenditure tax: The government would not manipulate the structure of savings
and investments, but by varying the tax rate on normal returns it could
approximate any position on the notional continuum between a pure income tax and
a pure expenditure tax. The latter is approximated if the (top) tax rate on
normal returns is reduced to zero, the former if this rate is aligned with the
top rate on labor and above-normal returns.
As this discussion shows, policy learning in OECD countries can be described as
a historical trial-and-error search for feasible compromises between two
ideal-types rather than a shift between incommensurable paradigms.
The alternative interpretation is fruitful in that it highlights oft-neglected
observations and questions: If policymakers and experts have found more
efficient ways of differentiating the tax burden, why has there been such a
strong downward trend in marginal tax rates on wages and above-normal returns (cf.
Wallerstein/Przeworski 1995)? Why has the Italian dual income tax recently been
abolished rather than exported to other countries? The next section deals with
these types of questions.
The "spillover problem" and the role of tax competition
One of the main purposes of corporation tax is to function as a "safeguard" for
personal income tax, and this function is best fulfilled if the corporation tax
rate is not below the top rate on wages, or at least not too far below. However,
corporate tax competition is strongest with respect to statutory tax rates and
the rates on above-normal returns. As a result, tax competition tends to spill
over into personal income taxation, increasing the costs of high marginal tax
rates on less mobile capital income and wages. It has thus contributed to the
downward trend in personal income tax rates.
Let me first discuss tax competition. Political scientists have generally been
skeptical about tax-competition explanations of corporate tax rate cuts, on both
empirical and theoretical grounds. The empirical objection is that corporate tax
revenue (as % of GDP) has not fallen in most OECD countries (for example,
Garrett 1998a). The theoretical objection is that if countries compete on tax
rates on normal returns to capital, income taxation in open economies is not
fundamentally different than in closed ones. Expenditure taxes can still
reconcile equity and efficiency goals (Wallerstein/Przeworski 1995). Yet both
arguments, while correct, provide no challenge to the tax-competition
explanation proposed here. First, corporate tax revenue as a percentage of GDP
tells us little about competitive dynamics since it is influenced by several
factors other than statutory corporate tax provisions (share of profits in GDP,
relative size of the corporate sector). Second, there is evidence that countries
have competed mainly on statutory tax rates as well as effective rates on
above-normal profits (that is, inframarginal investments).
The first type of competition is mainly driven by the profit-shifting behavior
of multinational firms (Hines Jr. 1999; Haufler/Schjelderup 2000), the second by
competition for highly profitable investments among such firms (Bond 2000;
Devereux/Griffith/Klemm 2002). Since very profitable investments, by definition,
generate a high share of above-normal returns, tax allowances are less important.
Revenue-neutral reforms that cut both statutory tax rates and allowances
therefore shift the tax burden from more profitable to less profitable
businesses. Such reforms imply higher taxation of normal returns of domestic
firms, but in small open economies this negative effect may be mitigated by the
lower statutory rates on inward investment (Bond 2000: 173).
There is robust evidence that the strong downward trend in corporate tax rates
since the mid-1980s has been driven to a large extent by tax competition (for an
overview see Griffith/Klemm 2004). The simplest way to see this is to look at
the relationship between tax rates and country size. Economic theory predicts
that small countries will have lower tax rates in equilibrium because their own
capital stock/tax base is small relative to potential inflows (Bucovetsky 1991).
The data pattern presented in Figure 2 is in line with this prediction. The
downward trend in corporate tax rates in advanced OECD countries is associated
with an increasing correlation between tax rates and country size (ln[population]).
Hence while the absolute convergence of corporate tax rates as measured by the
standard deviation has been moderate, the obvious pattern of conditional
convergence (Sala-i-Martin 1996) is consistent with the theory of asymmetric tax
rates and their correlation with country size in 21 advanced OECD
Notes: The sample includes the cases included in Table 1 except Luxembourg. The
average is unweighted. Correlation coefficient = Pearson's r.
Sources: Population: World Bank (2003), corporate tax rates: Ganghof (2006
Qualitative evidence supports this conclusion. In many countries, policymakers
clearly would have maintained generous investment incentives for businesses (for
example, accelerated depreciation rules) had it not been for competitive
pressures on statutory corporate tax rates. Examples include Australia, Canada,
Denmark, and Germany (Ganghof 2006 forthcoming). In addition, there is evidence
that tax competition has created a bias against pure expenditure taxes,
as these would require higher statutory corporate rates in order to achieve some
given revenue level. In the United Kingdom this was one major argument against
expenditure taxes (Isaac 1997: 313). Croatia did operate a pure direct
expenditure tax between 1994 and 2001, but eventually decided to abandon this
system and lower the corporate rate from 35% to 20% (Keen/King 2003).
How can falling corporate tax rates create problems for income taxation as a
whole? Consider first why there are corporation taxes at all. After all, every
tax falls eventually on individuals. One main reason is that corporation tax
functions as a safeguard for personal income tax (Mintz 1995). If the retained
profits of incorporated businesses went untaxed, much personal income tax could
be avoided. For the same reason, tax designers and international institutions
strongly recommend that "personal and corporate income taxes have the same top
marginal rate" (Stotsky 1995: 282) or "not differ materially" (Tanzi/Zee 2000:
310). If the corporate tax rate is much below the top personal rate, high-income
taxpayers have a large incentive to shift their income into the corporate sector
(cf. Gordon/Slemrod 2000; Fuest/Weichenrieder 2002; Wallerstein/Przeworski 1995:
These recommendations do not distinguish between income and expenditure taxes
because taxation of above-normal returns is sufficient to complement taxation of
wages. A pure wage tax, by contrast, which also exempts above-normal returns,
does create significant problems for tax administrations:
Under a wage tax, relabeling your labor compensation as capital income would be
an easy avenue to escape taxation altogether. The difficulty of distinguishing
what is labor income and what is capital income is an important reason that a
pure wage tax would end up being highly inequitable and costly to enforce. (Slemrod/Bakija
This implies that enforcement costs also tend to be high when the marginal tax
rates on both normal and above-normal returns are well below the marginal rates
Hence by putting pressure on the taxation of above-normal returns at the
corporate level, tax competition makes it more difficult for corporation tax to
fulfill its safeguarding function for personal income taxation. Tax competition
raises the costs of imposing high marginal tax rates on wages and on the
above-normal returns of unincorporated businesses. The tax reform experiments in
Croatia and Italy exemplify this problem. When Croatia introduced a progressive
expenditure tax in 1994, the Ministry of Finance proposed a corporate rate of
35% to align it with the top rate on personal income. Tax competition motivated
the parliamentary majority to set it at 25%, but, as expected, the gap between
the two top rates led to significant tax avoidance problems (Rose/Wiswesser
1998: 272). The corporate rate was raised to 35%, but given tax competition it
was only a question of time until it had to fall again. When it did (to 20% in
2001), the expenditure tax approach was abolished altogether (Keen/King 2003).
The Italian experience is similar. The Italian-style dual income tax had itself
been a concession to tax competition: the tax rate on normal profits was set at
19% rather than zero - as under a pure expenditure tax - so that the tax rate on
above-normal returns could be lower, given a revenue target (Bordignon/Giannini/Panteghini
2001). As in Croatia, however, the preferential taxation of normal profits was
recently abolished in return for a slightly lower corporate tax rate.
It is not surprising, therefore, that Nordic-style dual income taxes do not tax
above-normal returns more highly at the corporate level. The consequence, though,
is that high-income earners, especially the owners of unincorporated businesses
and small corporations, have greater incentives to transform wages into capital
income (Hagen/Sorensen 1998). The resulting tax avoidance increases the
administrative and political costs of differentiated income taxation and thus
creates additional pressure to reduce marginal rates on personal income.
I do not intend to exaggerate the problem. There surely are ways to deal with a
large tax rate gap between corporate and personal income taxation. Perhaps the
most systematic one is to retain the basic logic of the Italian-style dual
income tax but to shift the taxation of above-normal returns from the level of
the corporation to the level of the shareholder (Sørensen 2006 forthcoming). The
idea is to include in the tax base of the progressive income tax the difference
between realized income from shares (dividends and realized capital gains) and
normal returns. Norway recently introduced a variant of such a system (Ganghof
2006 forthcoming: chapter 6). The problem, however, is that taxing shareholders'
above-normal returns leads to great administrative complexity. In Norway a
centralized, computerized register of all Norwegian personal shareholders was
created to keep track of the (adequately) adjusted values of shares and of the
utilization of tax allowances for normal profits. Moreover, the Norwegian model
nevertheless required further cuts in the top marginal tax rates.
Hence while there are ways of making a large tax rate gap between corporate and
personal taxation work, they are costly. As a result, the decks seem to be
stacked in favor of flatter income taxes. We can already see this empirically.
In Eastern European transition economies flat income taxes rather than
progressive expenditure taxes are spreading. Such taxes were adopted in Estonia
(1994: 26%, to be reduced to 20% in 2007), Latvia (1995: 25%), Russia (2001:
13%), Serbia (2003: 14%), Ukraine (2004: 13%), Slovakia (2004: 19%), Romania
(2005: 16%), and Georgia (2005: 12%) and are likely to be adopted in further
countries, such as Poland.
While flat income taxes are now also much more seriously discussed in advanced
OECD countries such as Germany and Italy, the higher overall tax burdens on
labor have mitigated against a radical flattening of income tax schedules (see
section 3). Nevertheless, there is evidence that corporate tax cuts have
contributed to cuts in the marginal income tax rates of individuals. Elsewhere I
present case-study evidence for a number of countries, including Australia,
Denmark, Germany, and New Zealand (Ganghof 2006 forthcoming; Ganghof/Eccleston
2004). Here I perform a simple cross-sectional regression analysis.
If low corporate tax rates increase the costs of maintaining high personal rates,
we can expect a systematic positive association between corporate and top
personal tax rates (cf. Slemrod 2004). I test this hypothesis with a simple OLS
regression for the year 2004. The sample (N = 38) includes all OECD members and
all actual and prospective EU members (that is, including Bulgaria, Romania, and
rates are for general government and are taken from a variety of sources
(Ganghof 2006 forthcoming: Appendix). I include two important control variables
(cf. Slemrod 2004). Total taxation as a percentage of GDP serves as a proxy of
countries' overall average effective tax rate on wages. This average tax burden
influences personal income tax rates either directly (if high total taxes go
together with high income taxes) or indirectly - if income taxes are low but
serve as a progressive counterweight to the extensive reliance on payroll taxes
and indirect consumption taxes. The variable also accounts for the possibility
that low overall taxation may be a common cause of low corporate and personal
tax rates. The second control variable is GDP per capita.
Notes: OLS estimation,
standard errors in parentheses. All taxation data are for general gov-ernment.
The tax rate data are for 2004, the other data for the latest available year.
Sources: Tax revenue is taken from Eurostat (2004) for EU members, from OECD
(2004b) for the remaining OECD countries, and from IMF country reports for the
rest. GDP per capita is taken from World Bank (2003).
The simple model performs surprisingly well. It accounts for around 60% of the
variation in top income tax rates - 70% if Luxembourg is dropped as an outlier -
and suggests that higher (lower) corporate tax rates are systematically
associated with higher (lower) top personal tax rates. Hence if we treat
corporate tax rates as the exogenous variable - a reasonable assumption given
strong tax competition - the results suggest that lower corporate tax rates do
indeed tend to pull down personal rates on high wages and above-normal capital
The point of this preliminary analysis is not to enter into the grand debates
about the domestic effects of globalization, but to highlight two important
observations. First, the politics of income tax structure has an international
dimension. EU member states that want to maintain high marginal income tax rates
on wages and above-normal returns would probably benefit greatly from a minimum
EU corporate tax rate (Ganghof/Genschel 2006 forthcoming). Second, tax policy
instruments like progressive expenditure taxes look more attractive "on paper" (Przeworski
1999: 43) than in real-world open economies. This contributes to limiting the
differences between the policy preferences of left-wing and right-wing parties
at the domestic level.
The domestic politics of differentiated income taxation
It follows from the above discussion that domestic political struggles about tax
policy are to a large extent struggles about the progressivity of wage
taxation in the tax system at large. In this section I focus on these
struggles and argue that straightforward left-right differences exist but that
they are often not very large. By extension, the importance of veto institutions
is generally also reduced. But there are exceptions. I will focus on Germany,
where different powerful veto points and tax competition interacted in
ways that forced a Social Democratic-led government to pursue regressive tax
Left-wing parties are more likely than right-wing parties to prefer more
progressive income taxation of wages even if this implies a large tax rate gap
between corporate and personal taxation and thus significant administrative
costs. This preference is, in part, based on efficiency considerations:
Progressive income taxes generally fall less heavily on low-wage earners than
payroll taxes and indirect consumption taxes, which may mitigate the employment
problems of the low-skilled (Scharpf 2000; Kemmerling 2005). Hence progressive
and differentiated income taxation is one major way of targeting tax reductions
to the low-skilled rather than cutting wage taxes across the board, which is
costly in revenue terms. Right-wing parties, in contrast, are more likely to
choose lower and flatter income taxes. As a result, it is possible to explain
many features of income tax reforms in terms of the interaction of institutional
power and partisan preferences. Elsewhere I provide systematic quantitative and
qualitative evidence for this claim (Ganghof 2005a, 2006 forthcoming). Here I
only give examples, focusing on changes in top personal income tax rates.
In countries like France, Greece, New Zealand, and the United States, the
alternation of right-leaning and left-leaning governments was systematically
associated with the fall and rise of top personal income tax rates. In countries
like Norway or Sweden, veto players - or merely influential players - outside
the government also played an important role. Swedish Social Democrats
implemented deeper cuts in the top personal tax rate when they needed the
support of the Liberals in the late 1980s/early 1990s, but later increased the
top rate again to 57% with the support of the left-wing parties in parliament.
Similarly, the Norwegian Labor Party used its power as an opposition party
vis-à-vis a small centrist minority government to increase the top personal rate
to 55% in 2000.
As already noted, however, the differences between the policy preferences of
left-leaning and right-leaning governments are typically not very large. In
Finland, for instance, the economic constraints of the 1990s were so tight that
the policy preferences of left-wing and right-wing parties converged strongly.
Marginal tax rates on capital had to be low and tax relief for wages was
desirable, but the combination of budget deficits, population ageing, and
unemployment required that the effective tax burden on capital increase
significantly and that tax relief for wages be targeted on low wages - through
lower income tax rates at the bottom and an earned-income tax credit. Similarly,
centre-right coalitions in Austria and the Netherlands accepted high marginal
tax rates on wages in order to maintain a significant degree of wage tax
progressivity (compare Table 1). In all of these cases there is little evidence
that the relevant right-wing actors deliberately pursued "inefficient" tax
policies in order to reduce the level of taxation.
One case in which the specific interaction of economic, partisan, and
institutional factors did have a rather drastic effect on legislative outcomes
is Germany. When a left-leaning coalition of Social Democrats and Greens took
office in 1998, Germany's total tax level was not particularly high (around 42%
of GDP) but its tax structure was rather inefficient. The statutory corporate
tax rate, like other marginal rates on capital, was extremely high by
international standards (around 57% including local taxes, compare Figure 2),
and there was broad partisan consensus that it had to fall quickly and
drastically. On the other hand, the top personal rate was only slightly above
the EU average (57% as compared to 53%) and the average effective tax burden on
capital was rather low (Table 1). This was partly due to the country's strong
reliance on regressive payroll taxes (18% of GDP), which also financed
expenditures unrelated to social insurance and which caused a heavy tax burden
on lower wages. Finally, there was an urgent need to put the public finances in
One would have expected a Social Democratic-led government to deal with this
situation by following a strategy similar to that of governments in Scandinavia
or the Netherlands: First, implement some kind of differentiated or dual income
tax, which greatly reduces marginal tax rates on capital while maintaining or
even increasing the revenue raised from capital and higher wages. Second, reduce
the tax burden on low wages, for example, by introducing a sizeable basic
allowance into the system of payroll taxes (cf. Scharpf 2001). The last thing
one would have expected was large general cuts in the marginal tax rate
on very high incomes - capital and labor incomes - and a large reduction in an
already moderate income tax burden. Yet this is precisely what happened. The
government cut income tax rates across the board and tried hard to limit the tax
rate gap between corporate and personal taxation. Hence the top personal tax
rate fell to 44% in 2005, which is low by international comparison, the
corporate rate to 39%, which is still high by international comparison. The
government puts the overall net tax reduction achieved by its tax reforms
between 1998 and 2005 at almost €60 billion. This amounts to around 3% of GDP
and is roughly equal to Germany's structural budget deficits (in 2002).
A critical contributory cause of this puzzling outcome is Germany's combination
of strong bicameralism and a far-reaching constitutionalisation of tax policy
(Ganghof 2006 forthcoming). The initial tax reform strategy of the Social
Democrats was indeed quite similar to the one sketched above, but the government
had to deal with two very powerful veto players. The first is the upper house (Bundesrat),
in which the government lost its majority soon after getting into office. This
gave veto power to an opposition that firmly rejected the idea of differentiated
income taxes. That the oppositional Christian Democrats adopted this stance may
seem surprising, given its own welfare state legacy, as well as the pragmatic
policies of Christian and Conservative parties in other countries. It is
explained to a large extent by the existence of a second powerful veto player:
the Federal Constitutional Court. Under the intellectual leadership of former
judge and tax expert Paul Kirchhof, the Court deliberately removed central
aspects of tax policy from the realm of democratic politics.
In respect of income taxation, it was widely and plausibly believed that the
Court would not accept any significant gap between the corporate tax rate and
the top rate on personal income. In fact, after Professor Kirchhof had left the
Court in 1999, he immediately became Germany's leading proponent of a 25% flat
tax and explicitly claimed that different tax burdens on labor and capital are,
in principle, unconstitutional.
The implicit veto threat of the Constitutional Court was backed by the Federal
Fiscal Court in April 1999. It considered one type of differentiated income tax
to be unconstitutional, one on which the Red/Green government had based its
initial tax reform step: To cut the corporate rate without discriminating
against unincorporated businesses, these businesses got a cut in the top
personal tax rate as well; the top rate for other personal income, notably wages,
remained unchanged. After the Fiscal Court's decision, this simple model of
differentiated income taxation was dead, and in the end the government could
only achieve bicameral agreement on a significant reduction of the corporate tax
rate by cutting income taxes across the board (for more detail see Ganghof 2006
forthcoming: chapter 7).
In sum, the puzzling outcome of German tax reforms is explained by the complex
interaction of economic, partisan, and institutional factors: Strong tax
competition rendered the legal status quo so unattractive that Social Democrats
could not simply veto any policy change, as they had done during their time in
opposition. At the same time, their preferred reform strategy - differentiated,
revenue-neutral tax rate cuts - was unacceptable to the relevant veto players.
The result was not only neoliberalism by surprise (Stokes 2001) but also
neoliberalism by default.
Large welfare states seem to require moderate capital taxation. Moderation can
be achieved in two ways: shifting the tax burden onto payroll taxes and indirect
consumption taxes (low income taxation) or privileging capital income
within the income tax (differentiated income taxation). One elegant
variant of the second option is a direct expenditure tax that fully exempts the
normal return on capital, while taxing above-normal profits together with wages.
Compared with low income taxation, differentiated income taxation tends to lead
to more progressive taxation of wages, which may have both equity and efficiency
advantages. However, large tax rate differentials between different types of
incomes are associated with economic, administrative, and political costs.
High-income taxpayers can find ways to transform high taxed incomes into low
taxed incomes, and, partly as a result, voters may resent visible forms of
differentiated income taxation. Low capital income taxation can thus spill over
into wage taxation and pull down marginal tax rates at the upper end of the wage
scale. Tax competition exacerbates this problem by putting pressure on corporate
tax rates on above-normal profits. It thus undermines governments’ ability to
use corporation tax as a safeguard for progressive personal income tax.
At national level, political conflicts are to a large extent about the overall
progressivity of wage taxation. Left parties prefer more progressive taxation,
even if this increases administrative complexity and implies visible
discrimination against wages within the income tax. Left-right differences,
however, are often not large, as governments face similar constraints. There is
little evidence that right-wing governments or veto players deliberately opt for
inefficient tax structures on order to create downward pressure on the overall
tax level. As the German case shows, however, particular interactions of
economic, partisan, and institutional factors may lock countries into rather
inefficient tax structures, at least temporarily.
Aaron, Henry J./Harvey Galper/Joseph A. Pechman (eds.), 1988: Uneasy Compromise.
Problems of a Hybrid Income-Consumption Tax. Washington, DC: Brookings.
Avi-Yonah, Reuven S., 2004: Risk, Rents, and Regressivity: Why the United States
Need Both an Income Tax and a VAT. In: Tax Notes, 1651-1666.
Becker, Gary S./Casey B. Mulligan, 2003: Deadweight Costs and the Size of
Government. In: Journal of Law and Economics 46, 293-340.
Bird, Richard/Eric M. Zolt, 2005: Redistribution via Taxation: The Limited Role
of the Personal Income Tax in Developing Countries. Toronto: University of
Boadway, Robin/Neil Bruce, 1984: A General Proposition on the Design of a
Neutral Business Tax. In: Journal of Public Economics 24, 231-239.
Bond, Stephen R., 2000: Levelling up or Levelling Down? Some Reflections on the
ACE and CBIT Proposals, and the Future of the Corporate Tax Base. In: Sijbren
Cnossen (ed.), Taxing Capital Income in the European Union. Oxford: Oxford
University Press, 161-179.
Bordignon, Massimo/Silvia Giannini/Paolo Panteghini, 2001: Reforming Business
Taxation: Lessons from Italy? In: International Tax and Public Finance 8,
Bucovetsky, Sam, 1991: Asymmetric Tax Competition. In: Journal of Urban
Economics 30, 167-181.
Carey, David/Josette Rabesona, 2002: Tax Ratios on Labour and Capital Income and
on Consumption. In: OECD Economic Studies 35, 129-174.
Cnossen, Sijbren, 2000: Taxing Capital Income in the Nordic Countries: A Model
for the European Union? In: Sijbren Cnossen (ed.), Taxing Capital Income in the
European Union. Oxford: Oxford University Press, 180-213.
Cusack, Thomas R./Pablo Beramendi, 2003: Taxing Work: Some Political and
Economic Aspects of Labour Income Taxation. WZB Discussion Paper SP II 2003 – 17
(October 2003). Berlin: Wissenschaftszentrum für Sozialforschung.
De Long, J. Bradford/Lawrence H. Summers, 1991: Equipment Investment and
Economic Growth. In: Quarterly Journal of Economics 106, 445-502.
Devereux, Michael P./Rachel Griffith/Alexander Klemm, 2002: Corporate Income Tax
Reforms and International Tax Competition. In: Economic Policy 17, 451-495.
Devereux, Michael P./Ben Lockwood/Michela Redoano, 2004: Do Countries Compete
over Corporate Tax Rates? Unpublished manuscript. Coventry: University of
Ervik, Rune, 2000: The Hidden Welfare State in Comparative Perspective: Tax
Expenditures and Social Policy in Eight Countries. Unpublished doctoral
dissertation. Bergen: University of Bergen, Department of Comparative Politics.
Eurostat, 2004: Structures of the Taxation System in the European Union. Data
1995-2002. Luxembourg: Eurostat.
Frank, Robert H., 1999: Luxury Fever. New York: Free Press.
Fuest, Clemens/Alfons Weichenrieder, 2002: Tax Competition and Profit Shifting:
On the Relationship between Personal and Corporate Tax Rates. In: Ifo-Studien
Ganghof, Steffen/Richard Eccleston, 2004: Globalisation and the Dilemmas of
Income Taxation in Australia. In: Australian Journal of Political Science 39,
Ganghof, Steffen, 2005a: The Conditional Importance of Veto Institutions:
Economic Constraints and Top Marginal Income Tax Rates in Advanced OECD
Countries. Unpublished manuscript. Cologne: Max Planck Institute for the Study
Ganghof, Steffen, 2005b: Globalisation, Tax Reform Ideals and Social Policy
Financing. In: Global Social Policy 5, 77-95.
Ganghof, Steffen, 2006: The Politics of Income Taxation. Colchester: ECPR Press,
Ganghof, Steffen/Richard Eccleston, 2004: Globalisation and the Dilemmas of
Income Taxation in Australia. In: Australian Journal of Political Science 39,
Ganghof, Steffen/Philipp Genschel, 2006: Tax Policy in an Enlarging EU: Is there
a Case for a Minimum Company Tax Rate? MPIfG Discussion Paper 06/1. Cologne: Max
Planck Institute for the Study of Societies, forthcoming.
Ganghof, Steffen, 2007: The Political Economy of High Income Taxation: Capital
Taxation, Path Dependence and Political Institutions in Denmark. In: Comparative
Political Studies, forthcoming.
Garrett, Geoffrey, 1998a: Partisan Politics in the Global Economy. Cambridge:
Cambridge University Press.
Garrett, Geoffrey, 1998b: Shrinking States? Globalization and National Autonomy
in the OECD. In: Oxford Development Studies 26, 71-97.
Garrett, Geoffrey/Peter Lange, 1991: Political Responses to Interdependence:
What’s Left for the Left? In: International Organization 45, 539-564.
Gordon, Roger H./Joel Slemrod, 1988: Do We Collect any Revenue from Taxing
Capital Income? In: Tax Policy and the Economy 2, 89-130.
Gordon, Roger H./Joel B. Slemrod, 2000: Are "Real" Responses to Taxes Simply
Income Shifting Between Corporate and Personal Tax Bases? In: Joel B. Slemrod
(ed.), Does Atlas Shrug? The Economics Consequences of Taxing the Rich. New
York: Russell Sage/Harvard University Press, 240-280.
Gordon, Roger/Laura Kalambokidis/Joel Slemrod, 2004: Do We Now Collect any
Revenue from Taxing Capital Income? In: Journal of Public Economics 88,
Griffith, Rachel/Alexander Klemm, 2004: What Has Been the Tax Competition
Experience of the Last 20 Years? Institute for Fiscal Studies, WP 04/05. London.
Hagen, Kare Peter/Peter Birch Sorensen, 1998: Taxation of Income from Small
Businesses: Taxation Principles and Tax Reforms in the Nordic Countries. In:
Peter Birch Sorensen (ed.), Tax Policy in the Nordic Countries. Houndmills:
Haufler, Andreas/Guttorm Schjelderup, 2000: Corporate Tax Systems and Cross
Country Profit Shifting. In: Oxford Economic Letters 52, 306-325.
Hettich, Walter/Stanley L. Winer, 1999: Democratic Choice and Taxation. A
Theoretical and Empirical Analysis. Cambridge: Cambridge University Press.
Hines Jr., James R., 1999: Lessons form Behavioral Responses to International
Taxation. In: National Tax Journal 52, 305-322.
Howard, Christopher, 1997: The Hidden Welfare State. Tax Expenditures and Social
Policy in the United States. Princeton: Princeton University Press.
IFS Capital Taxes Group, 1991: Equity for Companies: A Corporation Tax for the
1990s. IFS Commentary no. 26. London: Institute for Fiscal Studies.
Isaac, John, 1997: A Comment on the Viability of the Allowance for Corporate
Equity. In: Fiscal Studies 18, 303-318.
Kato, Junko, 2003: Regressive Taxation and the Welfare State. Path Dependence
and Policy Diffusion. Cambridge: Cambridge University Press.
Keen, Michael/John King, 2003: The Croatian Profit Tax: An ACE in Practice. In:
Manfred Rose (ed.), Integriertes Steuer- und Sozialsystem. Heidelberg: Manfred
Kemmerling, Achim, 2005: Tax Mixes, Welfare States and Employment: Tracking
Diverging Vulnerabilities. In: Journal of European Public Policy 12, 1-22.
Kirchhof, Paul, 2003: Der Grundrechtsschutz des Steuerpflichtigen: Zur
Rechtsprechung des Bundes-verfassungsgerichts im vergangenen Jahrzehnt. In:
Archiv des öffentlichen Rechts 128, 1-51.
Layard, Richard, 2005: Happiness. Lessons from a New Science. New York.
Lindert, Peter H., 2004: Growing Public: Social Spending and Economic Growth
since the Eighteenth Century. Volume One: The Story. Cambridge: Cambridge
Mintz, Jack M., 1995: Corporation Tax: A Survey. In: Fiscal Studies 16, 23-68.
Murphy, Liam/Thomas Nagel, 2002: The Myth of Ownership. Taxes and Justice.
Oxford: Oxford University Press.
OECD, 2004a: Employment Outlook. Paris: OECD.
OECD, 2004b: Statistical Compendium. OECD.
OECD, 2004c: Taxing Wages 2003-2004. Special Feature: Broadening the Definition
of the Average Worker. Paris: OECD.
Przeworski, Adam, 1999: Minimalist Conception of Democracy: A Defense. In: Ian
Shapiro/Casiano Hacker-Cordón (eds.), Democracy's Value. Cambridge: Cambridge
University Press, 23-55.
Przeworski, Adam/Michael Wallerstein, 1982: Democratic Capitalism at the
Crossroads. In: Democracy 2, 52-86.
Przeworski, Adam/Michael Wallerstein, 1988: The Structural Dependence of the
State on Capital. In: American Political Science Review 82, 11-29.
Røed, Knut/Steinar Strøm, 2002: Progressive Taxes and the Labour Market – Is the
Trade-Off between Equality and Efficiency Inevitable? In: Journal of Economic
Surveys 16, 77-110.
Rose, Manfred/Rolf Wiswesser, 1998: Tax Reform in Transition Economies:
Experiences from the Croation Tax Reform Process of the 1990s. In: Peter Birch
Sørensen (ed.), Public Finance in a Changing World. Houndsmills: Macmillan,
Sala-i-Martin, Xavier, 1996: The Classical Approach to Convergence Analysis. In:
The Economic Journal 106, 1019-1036.
Scharpf, Fritz W., 2000: The Viability of Advanced Welfare States in the
International Economy: Vulnerabilities and Options. In: Journal of European
Public Policy 7, 190-228.
Scharpf, Fritz W., 2001: Employment and the Welfare State. A Continental
Dilemma. In: Bernhard Ebbinghaus/Philip Manow (eds.), Comparing Welfare
Capitalism. Social Policy and Political Economy in Europe, Japan and in the USA.
London: Routledge, 270-283.
Slemrod, Joel, 1990: Optimal Taxation and Optimal Tax Systems. In: Journal of
Economic Perspectives 4, 157-178.
Slemrod, Joel, 1997: Deconstructing the Income Tax. In: American Economic Review
Slemrod, Joel, 2004: Are Corporate Tax Rates, or Countries, Converging? In:
Journal of Public Economics 88, 1169-1186.
Slemrod, Joel/Jon Bakija, 2004: Taxing Ourselves. A Citizen's Guide to the Great
Debate Over Tax Reform. 3rd. edition. Cambridge: MIT Press.
Sørensen, Peter B., 1998: Recent Innovations in Nordic Tax Policy: From the
Global Income Tax to the Dual Income Tax. In: Peter Birch Sørensen (ed.), Tax
Policy in the Nordic Countries. London: Macmillan, 1-27.
Sørensen, Peter Birch, 2006: Neutral Taxation of Shareholder Income. In:
International Tax and Public Finance, forthcoming.
Steinmo, Sven, 2003: The Evolution of Policy Ideas: Tax Policy in the 20th
Century. In: British Journal of Politics and International Relations 5, 206-236.
Stokes, Susan C., 2001: Mandates and Democracy. Neoliberalism by Surprise in
Latin America. Cambridge: Cambridge University Press.
Stotsky, Janet, 1995: Summary of IMF Policy Advice. In: Parthasarathi Shome
(ed.), Tax Policy Handbook. Washington, DC: International Monetary Fund,
Summers, Lawrence/Jonathan Gruber/Rodrigo Vergara, 1993: Taxation and the
Structure of Labor Markets: The Case of Corporatism. In: The Quarterly Journal
of Economics 108, 385-411.
Swank, Duane, 1998: Funding the Welfare State: Globalization and the Taxation of
Business in Advanced Market Economies. In: Political Studies 46, 671-692.
Swank, Duane/Sven Steinmo, 2002: The New Political Economy of Taxation in
Advanced Capitalist Democracies. In: American Journal of Political Science 46,
Tanzi, Vito/Howell Zee, 2000: Tax Policy for Emerging Markets: Developing
Countries. In: National Tax Journal 53, 299-322.
Wallerstein, Michael/Adam Przeworski, 1995: Capital Taxation with Open Borders.
In: Review of International Political Economy 2, 425-445.
Wilensky, Harold, 2002: Rich Democracies. Political Economy, Public Policy, and
Peformance. Berkeley: University of California Press.
Willoch, Kaare, 1995: What Can Other Countries Learn from the Scandinavian
Experience? II. In: Nordic Journal of Political Economy 22, 177-182.
World Bank, 2003: World Development Indicators, CD ROM. Washington, DC: The
I use the terms "expenditure tax" and "consumption tax" interchangeably.
For the purposes of this paper I ignore compensation for inflation and returns
to risk-taking, which need not be taxed under either income or consumption taxes.
For critical discussions of these and other arguments see Murphy/Nagel (2002)
and Avi-Yonah (2004).
Most data on average effective tax burdens on capital, labor, and consumption
available for larger sets of countries, including that of Carey/Rabesona (2002),
does not adequately reflect many forms of tax relief for capital income This can
lead to highly misleading data patterns. The Eurostat data seem more reliable
because Eurostat has access to national sources on the breakdown of personal
income tax revenue into its capital and labor components. These data are only
available for EU-15 member states and the period 1995-2002. Since the Eurostat
estimates are at least roughly comparable to those of Carey/Rabesona, however, I
combine the Eurostat data with these OECD data (for the period 1995-2000).
Wage coordination scores (for the period 1995-2000) are taken from OECD (2004a:
151). The estimated equation (OLS) is this: CapitalTax/(Labor+ConsumptionTax)
= 135.42 - 6.87WageCoordination - 1.81TotalTaxation. Adj.
R-squared = .77.
Here and throughout I assume that payroll taxes and indirect consumption taxes
are largely shifted backwards onto wages. Note also that the arguments of
Wilensky and Kato are partly based on the visibility of income and
property taxes. I will not comment on this mechanism here.
For a more detailed analysis of tradeoffs and the resulting conceptions of
policy change, see Ganghof (2006 forthcoming: chapters 3 and 4).
Similar estimates for the US are provided by Gordon/Slemrod (1988) and Gordon et
The DIT model has influenced tax reforms in many countries, although only a few
countries have so far implemented this model consistently (Ganghof 2005b). The
model is now also advanced as the most adequate model of pragmatic tax reform
for developing countries (Bird/Zolt 2005).
The underlying model of expenditure taxation is of a "pre-paid" kind. Rather
than exempting savings but taxing returns (post-paid or cash-flow expenditure
tax), savings are taxed but normal returns are exempt. On the resulting
alternative to cash-flow taxes see Boadway/Bruce (1984), as well as IFS Capital
Taxes Group (1991).
Some studies also find evidence for competition on effective tax rates on normal
profits (that is, "effective marginal tax rates") - see especially Devereux et
Ganghof (2005a) uses regression analysis to investigate the relationship between
country size and tax rates; Ganghof (2005b) shows that the relationship between
tax rates and country size also holds for a much larger sample of countries.
Note that in some of these countries the corporate tax rate is higher than the
flat tax rate on personal income.
Extending the sample is difficult because standard data sources such as World
Bank (2003) only report tax rate and revenue data for central government
(cf. Slemrod 2004).
In the English summary to a recent article, Kirchhof (2003: 50) states: "The democratic hope that the parliamentary process - the decision-making of the
representatives of all taxpayers - will guarantee a moderate and consistent tax
law has not been fulfilled. For this reason, the fundamental rights of taxpayers
in Germany are increasingly becoming the primary measure of legislation."
In autumn 2005 Kirchhof joined the election campaign of the Christian Democrats
as Shadow Finance Minister. His flat-tax proposal became one of the crucial
Copyright © 2006 Steffen Ganghof
No part of this publication may be
reproduced or transmitted without permission in writing from the author.
Jegliche Vervielfältigung und Verbreitung, auch auszugsweise, bedarf der
Zustimmung des Autors.
MPI für Gesellschaftsforschung,
Paulstr. 3, 50676 Köln, Germany