Abstract
This article argues that the “fiscal trilemma”—the tension between sufficient revenue and long-run fiscal sustainability on one side; medium- to long-run growth opportunities on the other; and progressivity and avoidance of inequality on the third side—is very much present in Denmark. We provide an overview of main features of the Danish tax system; the state of fiscal balances in the short and longer run; growth and productivity developments; and the extent of inequality in the Danish society. Thereafter, we go into details with reforms and changes in the tax system over the last twenty-five years and diagnose main problem areas. We argue that the tax reform carried in Parliament on September 13, 2012, does not pass a “trilemma test”; that is, it is in conflict with at least one of the three key considerations. Finally, we put forth reform elements that should improve the system.
The Danish Economy and its Recent History
On the surface, the Danish economy looks rather different from the US economy. It covers a society that arguably is more homogeneous in nature. Unlike the United States, Denmark has rather few natural resources, so must base its income generation on the fruits of human capital. The income distribution is among the most compressed in the world, and the welfare state is well developed, receiving strong backing from the population. 1 Public services and redistribution of income are in demand at the expense of private consumption opportunities. Although the trend is away from regulation and public sector involvement in the economy, there is still a marked difference to conditions in the United States. Finally, as a small, open economy, Denmark is under fierce competitive pressure from the rest of the world, not least its European neighbors. As a result, degrees of freedom in economic policy are becoming scarce. In addition, since Denmark is a member of the European Union (EU), key decisions in many policy areas must be taken at the EU level.
Table 1 comprises a few key variables for Denmark, the EU (average), and the United States. Denmark’s gross domestic product (GDP) per capita is significantly above the average of EU, yet below that of the United States. Average GDP growth has been relatively low in Denmark especially after 2001, whereas the rate of inflation has been quite similar over the last decade. Despite the ongoing crises, Denmark’s rate of unemployment is still low in an international perspective. Finally, the average EU country is much more open than the United States, and Denmark belongs to the more open countries in the EU.
Key Figures of the Danish Economy Compared to the United States and European Union, 2001 and 2010.
Source: Eurostat and stats.oecd.org.
Note: GDP = gross domestic product; HICP = harmonized indices of consumer prices; PPP = purchasing power parity.
a Note, that the calculation for the European Union in 2001 is only based upon data from 1995 to 2001.
b Harmonized indices of consumer prices, annual average rate of change. c Openness defined as (Export + import)/(2 × GDP). For the EU, it is the sum of countries’ exports and imports divided by twice European Union’s GDP.
Figure 1 zooms in on the growth experiences of Denmark, the eurozone, and the United States since 2000. In the beginning of the period, Denmark followed the eurozone, while the United States grew a bit faster. Since the outbreak of the financial crisis in 2008, Denmark has experienced a bigger fall in GDP than the other two regions.
Gross domestic product (GDP) in Denmark, United States, and the eurozone, 2000–2012. Source: Economic Councils.
The Danish economy was in an admirable position a decade ago. The growth rate, albeit not huge, was significantly positive; fiscal policy was close to sustainable with a current surplus; public indebtedness was moderate; the current account of the balance of payments had a surplus; and unemployment was heading down toward almost an all-time low. It was in fact an ideal time to undertake sensible adjustments of structural economic policy. Prime areas for reform would have been the labor market (strengthening the “flexicurity” 2 system), early retirement regulations, rules for welfare payments and publicly supported employment, and aspects of the tax system.
The general view among Danish economists is, though, that over the next decade too little was accomplished. 3 In particular, fiscal policy did not respond well to a booming economy, being in periods too lax, and reforms in early retirement and other areas were postponed. Unfortunately, the boom was coming to an end exactly when the financial crisis struck in 2008, causing a major dip in growth (GDP fell by more than 6 percent in 2009), from which the economy is still trying to recover. Furthermore, the acute attention that had to be devoted to the financial system stifled economic policy in other areas. In many ways, the last decade represents a phase of missed opportunities in terms of reforms and policy adjustments.
In fairness, the government and other parties in the parliament did enter a Welfare Agreement in 2006, entailing a gradual increase in the early retirement age as well as the ordinary retirement age (from 2019 to 2022) plus (after 2025) indexation of retirement ages to expected remaining lifetimes for a sixty-year-old. However, this agreement was strengthened in 2011 by an early retirement reform which moves the increase in early retirement age forward to 2014 and cuts the early retirement period from five to three years. 4
The Danish Tax System—Main Revenue Components
Tax policy is the responsibility of individual member states in the EU. That said, several restrictions are imposed at the EU level (for instance, concerning minimum rates for the value-added tax [VAT] and for certain excises). On top of that, international mobility of goods, production factors, and corporate income create a strongly competitive atmosphere for tax policy, a point to which we shall return.
The key features of the Danish tax system are listed in table 2. Table 2 shows the relative importance of the various revenue generators and also compares the Danish tax system with those in the other EU countries. From the table it is seen that one-half of total tax revenue comes from personal income taxation (including the personal gross income tax called labor market contributions). It is worth stressing that taxation of labor inputs in production almost exclusively takes the form of personal income tax. There are almost no social security contributions at the firm or personal level, and there is no payroll tax. This is one aspect in which the Danish tax system is unique. 5
Key Figures of the Danish Tax System, 2010.
Source: Eurostat.
Note: GDP = gross domestic product; VAT = value-added tax.
a Ranking within the twenty-seven member countries. See the publication for the computation of implicit tax rates.
b The mean USD-euro rate of 2010, has been employed to translate values to USD.
Another unique aspect of the system is its comparatively heavy reliance on the VAT and excises. Together, these indirect taxes comprise more than one-third of total revenue. The Danish VAT not only has the second-highest rate in the world (25 percent) but also has very broad coverage, few exemptions, and no reduced rates. Add to this fairly high excise taxes in the “sin” and “green” areas.
A third unique feature of the Danish tax system is the tax on the return to pension savings, be it occupational or individual pension arrangements. While at a low rate and invariably with volatile revenue, it does contribute some 3 percent of revenue. Most countries exempt the return to pension savings from tax, but Denmark instituted such a tax as early as 1983. It has since been altered on several occasions yet has remained a fixture of the tax system.
The corporate income tax is not a very important source of revenue—a point which is often overlooked in public debate. Only around 7-8 percent of revenue derives from taxing corporate income, and the rate, currently 25 percent, has come down significantly from 50 percent in the 1980s. Still, the corporate tax probably constitutes an essential backstop in the tax system.
The remaining revenue stems from taxes on inheritances and land and the like. There is no wealth tax; it was abolished in 1997 after yielding only a tiny fraction (less than 0.20 percent) of tax revenue.
Overall, tax revenue as a fraction of GDP is higher in Denmark than in other EU countries (and much higher than in the United States). 6 Judging from the so-called implicit tax rates, consumption is especially heavily taxed. Employed labor only has a tax burden corresponding to the EU average. 7
Although inspired by the so-called dual income tax (DIT; see the section on Tax Policy: Reforms and Adjustments Since the 1980s), the exact computation of tax with its interweaving of taxes on personal income and personal capital income is quite complicated. Personal capital income here consists mainly of interest income from bank accounts, bond holdings, and so on. Income from share holdings in the form of dividends or capital gains is taxed in a separate, progressive system with two rates (common rates for dividends and capital gains). Hereby, double tax relief is conceded (the combination of corporation tax and the highest dividend/capital gains tax corresponds quite closely to the top marginal tax on personal capital income). Imputed rent on housing likewise is taxed in a separate, progressive system. That tax has since 2001 been one of the prime subjects of the “tax freeze” (which we discuss in the following).
The Fiscal Situation and Sustainability
Table 3 provides an overview of the structure of the public budget in Denmark in 2011. The expenditure and revenue sides of the general government budget are brought together in table 3, which contains figures for 2011, issued in August 2012 as a prelude to negotiations on the state budget for 2013. It clearly shows the importance of government consumption and income transfers on the expenditure side. Also, it draws attention to the fact that part of taxes on labor are not conventional (personal) income taxes but rather take the form of a gross tax on earnings (called labor market contributions in the table), which allows no deductions.
Main Items in the Public Finances, 2011.
Source: Danish Ministry of Finance: Budget Outlook, August 27, 2012.
a Personal income taxes include withholding taxes, tax on imputed income from owner occupied dwellings, specific taxes from households, tax on estates of deceased persons, and other personal taxes.
b Social contributions (unemployment insurance contributions, retirement contributions, etc.).
Over the last fifteen years or so a growing consensus has developed in Denmark, according to which it is important to view the conduct of fiscal policy not only relative to the current economic situation in the country, but also in relation to the sustainability of government finances in the medium to long run. It has indeed become customary to give an account of how proposed fiscal policy changes impact longer-term sustainability. Until the outbreak of the financial crisis, the Danish government’s strategy was to run fiscal surpluses of 1½ to 2½ percent of GDP, so as to prepare for the upcoming demographic transition and keep fiscal policy on a sustainable path.
The Economic Council 8 of Denmark has repeatedly computed the so-called sustainability indicators for fiscal policy. The last decade’s computations of this indicator are summarized in table 4. The sustainability indicator shows the necessary once-and-for-all adjustment in expenditure relative to GDP necessary to ensure that the public sector, given the rules for future expenditures, and given the current tax system, has the funds for honoring its obligations over time. 9 For a long period, the sustainability indicator was computed to be negative, around 1½ percent of GDP; that is, one and one-half percentage points’ worth of cuts in public expenditure were deemed necessary to ensure fiscal policy sustainability. The most recent computation, however, has ended up with a small, positive number, suggesting that at present fiscal policy is just on a sustainable path.
Size of the Fiscal Sustainability Problem.
Source: The Danish Economic Councils.
This puts Denmark in a unique position. According to the sustainability indicator (S2) computations of the European Commission, all other EU member states (except for a few Eastern European countries) have moderate to colossal sustainability problems in their fiscal policies.
Each time the sustainability indicator is computed, a number of underlying assumptions change relative to the preceding round. Policy may have changed, general economic conditions, including not least interest rates and unemployment levels, may have changed. Important reasons for the shift in the indicator surely are the reform of the early retirement system agreed upon in 2011, and the introduction of indexation of a number of excise taxes, which had previously been held fixed by the tax freeze. Unprecedented low interest rates and low foreseen growth rates (dulling the mechanisms of Baumol’s disease) may also have contributed to rendering fiscal policy more sustainable.
In 2011, gross public debt to GDP (the Maastricht definition) amounted to 47 percent of GDP, way below the eurozone average of more than 80 percent. Net public debt (including government deposits in the central bank, government-owned entities’ private bond holdings, government loans to private sector entities, etc.) is very close to zero.
On the surface, then, public finances look sound. However, an underlying problem of public spending growth repeatedly exceeding targets was dulled by the booming economy before the crisis. During the crisis, the public deficit went beyond the 3 percent limit in the growth and stability pact of the EU, prompting the EU to issue a request that Denmark improve the structural budget balance by 1½ percent from 2010 to 2013.
Economic Growth and Productivity in Denmark
This section describes Denmark’s recent history of growth in output and productivity and the apparent problems related to very slow development of productivity. The average growth rates of Danish output, labor productivity, and total factor productivity (TFP) during the period 1995 through 2007 are presented in table 5. Moreover, the growth performance for the EU and the United States are presented for comparison.
Contributions to Growth of Real Output in the Market Economy, Denmark, European Economy, and the United States, 1995–2007 (annual average growth rates, in percentage points).
Source: van Ark (2011).
Note: TFP = total factor productivity.
a “ICT” is information and communications technology.
b Data for the European Union excludes five member states of EU-15: Greece, Ireland, Luxembourg, Portugal, and Sweden.
c Data for the Unites States is based on old standard industrial classification.
It is evident that output growth has been parallel to that of the EU and around 2½ percent per year. Output growth in the United States has been faster and around 3½ percent per year. When it comes to productivity growth, it has been relatively low in Europe and miserable in Denmark since the mid-1990s. In Denmark, development in labor productivity is lagging behind the two other regions. The same pattern is evident for TFP growth that attains a negative value in Denmark.
Figure 2 illustrates the implications of different growth rates in labor productivity. For example, labor productivity has increased by almost 40 percent in the United States during the period 1995 through 2010, whereas it has only increased by a little more than 10 percent in Denmark during the period.
Gross domestic product (GDP) per working hour (total economy), 1995–2010. Source: Andersen and Spange (2012).
What is the explanation for the European productivity problem—and in particular the Danish problem? The weak productivity performance is actually a paradox and not easily explained. 10 It is not at all clear which factors have been responsible for this development. Potential candidates are (1) research and development (R&D), (2) educational attainment, and (3) investments in new technology, that is, information and communication technology (ICT). However, relative to the United States, research intensity and educational attainment among the population of working age have improved and, as such, have not contributed to the widening of the productivity gap (Madsen and Sørensen 2011). On the other hand, the contribution to productivity growth from ICT capital among the EU countries has been somewhat lower for Europe and may as such be part of the explanation for the European productivity paradox (van Ark 2011). 11
Denmark, however, actually seems to be performing well with respect to all three factors:
Investments in R&D: if the Danish investment level is low, this may potentially explain the productivity performance. However, the level has previously been low relative to comparable countries, but in recent years Denmark is one of the countries with the largest growth in knowledge capital (Economic Council 2010; Andersen and Spange 2012).
Level of education: if the Danish educational level is low this may potentially explain the productivity performance. However, the share of population in working age with a higher education in Denmark is in line with most comparable countries. An alternative explanation related to education focuses on whether the allocation of skilled workers between the private and the public sector is appropriate. This is motivated by the finding that skill-intensive firms with innovation activities experience higher growth rates than unskilled-intensive firms with innovation activities (Caroli and Van Reenen 2001; Junge, Severgnini, and Sørensen 2012). In other words, skilled workers employed in the private sector may be important for productivity growth. There are some indications of private sector employment of educated labor being relatively low (Economic Council 2010).
Investments in information technology (IT): if the Danish investments in IT are low, this may potentially explain the productivity performance. However, contribution to labor productivity from IT service per hour worked is relatively high in Denmark, which is presented in table 5. In other words, Danish IT investments have been relatively high and therefore cannot contribute in the explanation of the Danish productivity problem.
An additional potential contributing explanation for the productivity puzzle suggested by Lazear and Shaw (2011) is the compressed wage structures in many European economies. In the words of Lazear and Shaw, “perhaps the most important point that comes from the compensation literature is that pay compression is the enemy of productivity.” This is an avenue that may be fruitful to explore in greater detail to gain some insight into the poor productivity performance of European economies.
Finally, a few words are needed on the relationship between the productivity performance and the tax system. The tax system influences the incentives that households and businesses face, and thus their decisions on consumption, employment, investment, choice of education, and so on. Therefore, the tax system should affect the productivity level and growth. As mentioned above in the section on The Danish Tax System—Main Revenue Components, the corporate tax rate equals 25 percent today, which is in line with other European countries. Personal taxes in Denmark are at the high end internationally, though. According to Arnold et al. (2011) both personal taxes and corporate taxes have a dampening effect on productivity growth. The negative effects act—according to the authors—primarily via weaker growth in TFP. The rationale is that higher taxes reduce after-tax return on investment that can raise TFP, lowering the incentive to undertake such investments.
In this light, high personal income taxes may be part of the relatively weak Danish productivity performance. On the other hand, Gemmell, Kneller, and Sanz (2011) find that in practice fiscal policy changes in the Organisation for Economic Co-operation and Development (OECD) countries mean that persistent increases or decreases in growth rates are rare.
Summing up, while the weak Danish productivity performance after 1995 may potentially be explained by relatively low private sector employment of educated workers and relatively high personal income taxes, it is unlikely that these factors constitute the full explanation. This is because the Danish productivity performance before 1995 was much stronger, and because low private sector employment of educated workers and relatively high personal income taxes prevailed also during this period.
Inequality and Income Redistribution
Denmark has one of the most equal income distributions in the world. This holds true whether income inequality is measured on before- or after-tax income. Figure 3 shows the Gini coefficient for disposable income. As can be seen from the figure, income inequality in Denmark is below that of even the other Nordic countries, is somewhat smaller than in continental European countries, and is significantly smaller than in the Anglo-Saxon countries. Furthermore, Danish income inequality has remained low and almost unchanged, albeit with some business cycle dependence, over the last two–three decades (Economic Council 2011).
Gini coefficients for selected countries, 2007 or 2008. Source: The Danish Economic Council.
Computation of Gini coefficients reveals the workings of the expenditure and revenue sides of the public budget. In 2009, the Gini coefficient for “private income” (labor income, including self-employment income, plus capital income) was about forty-three; for gross income (including public transfers) it was thirty, and for disposable income finally twenty-six (against thirty-eight in the United States). 12 So obviously, public transfers—pensions, welfare payments, child benefits, contributions to unemployment benefits, and so on—play a major role in evening out consumption possibilities. The high and relatively progressive Danish income tax also contributes significantly to form an equal distribution of consumption possibilities.
The extent to which the different income tax instruments contribute to lowering inequality was illustrated by computations done in Economic Council (2001). 13 These computations show the change in the Gini coefficient by implementing changes resulting in a similar revenue increase. The tax that has the largest redistributive effect is—as expected—the top tax bracket rate. This is true for both the capital income tax and the labor income tax. Going downward in the list of redistributive taxes, the next places are occupied by the middle bracket income tax rate, and the imputed rent tax on housing.
Also other tax instruments besides income taxation have bearings on consumption possibilities. The VAT is levied on all consumption with almost no exemptions in Denmark. Consumption taxes tend to have a rather low impact in terms of equalizing consumption possibilities as consumption has a smaller variance across the income distribution than does income itself—in other words, poor individuals consume less than rich individuals, but not as much less as the difference in income would point to. 14 Furthermore, the VAT is a flat-rate tax and thus is levied in equal fashion on small and large consumption bundles.
Denmark has large excise taxes in particular for cars. The current rate is 105 percent for small cars and 180 percent for larger cars. This does affect the equality of consumption possibilities somewhat as more well-off households tend to have more than one car and also to buy larger cars. Excise taxes in the green and sin areas are not reducing inequality in consumptions possibilities as per capita consumption in these areas are only loosely related to income.
Finally, property taxes are levied on both land and dwellings. The tax on dwellings (imputed rent tax on housing) is progressive with properties valued at more than 3 million DKK (on May 8, 2013, one DKK corresponds to US$0.176) being taxed at a higher rate than properties valued below this threshold. Property taxes are thus reducing the inequality in consumption possibilities in Denmark, albeit not to a large degree. Furthermore, property taxes have been an important part of the 2000s tax freeze—leading to a large fall in the effective property value tax rate for high-valued properties—see the following.
Tax Policy: Reforms and Adjustments since the 1980s
In the postwar period, Denmark developed, like most other Western countries, an income tax system, in which comprehensive income (income from all sources, labor as well as capital income) in principle was subject to a progressive tax. 15 On account of mounting problems it was decided to reform the system drastically in the mid-1980s, with effect from 1987 onward. Instead of the comprehensive income tax, a so-called dual income tax (henceforth, DIT) was introduced. 16
The idea of the DIT is to subject labor income (the fruits of labor) and capital income to different tax schedules. Ideally, the labor income tax would be progressive in the DIT, while the capital income tax would be flat rate. The DIT thus would offer redistribution of (labor) income while, with modest capital income tax rates, not hampering saving and investment as much as a comprehensive income tax.
The version of the DIT introduced in Denmark was not a pure one, however, and through subsequent reforms in the 1990s (1994, 1999) the income tax has become a complicated blur of a dual and a comprehensive income tax. Now, four successively higher rates are applicable to personal net capital income (interest expenditures, respective interest income). Large, negative capital income is deducted at the lowest rate, while high positive capital income is taxed at the highest rate. The progression in (positive) personal capital income taxation is reminiscent of but below that of the labor income tax.
Figure 4 shows how personal income taxation (or taxation of labor income) and the associated progressivity have evolved over the years. The figure contains information from the years 1986, 2002, 2009, and following the reform that took effect in 2010. The four curves depicting marginal tax rates have been transformed to make them comparable—all incomes have been transformed into 2009 incomes. The introduction of the gross labor earnings tax, the so-called labor market contributions, in 1994, is clearly seen in the figure. Unlike 1986, the later years entail taxation of even very small incomes.
Marginal tax rates across incomes over time.
Since the mid-1980s, the following have been the major trends in tax policy changes:
Reduction in personal (labor) income tax rates. Before the reform in the mid-1980s, the top marginal tax rate on comprehensive income had climbed all the way up to 73 percent. Besides its likely dampening effect on labor supply, an obvious effect was to cause high-earnings households to borrow extensively. After all, they would only have to cover 27 percent of marginal interest payments. The introduction of the DIT led to a lowering of the top marginal labor income tax rate, and subsequent reforms in the 1990s and again in the 2000s have reduced the top rate to around 56 percent at present (figure 4), while at the same time broadening the income tax base by limiting deductions and eliminating loopholes. Inspired by experiences in other countries, a version of earned income tax credit has been introduced in 2004 in order to provide an extra incentive to take up work. The credit has since been raised on a couple of occasions.
Reduction in the corporate income tax rate (coupled with attempts at base broadening). In the beginning of the 1990s, the corporate income tax rate stood at 40 percent, having fallen from 50 percent in the 1980s. In the major reform in the mid-1980s, it was reduced markedly, and with reference to increased international tax competition in the corporate area it has been reduced frequently in the last decade, so that at present it stands at 25 percent. In order to trim the tax and recoup revenue, the tax base has been broadened (this also with inspiration from tax reforms in other countries).
Figure 5 contains the statutory corporate income tax rate in Denmark since 1987 and thus the remarkable halving of the rate in that time span. The figure also includes the (unweighted) average corporate tax rate for the EU countries. Denmark’s corporate tax has in fact followed the average EU rate quite closely and is today slightly above it.
Adjustments in excise tax rates. Throughout especially the 1990s and also afterward, some attention has been devoted to the phenomenon of cross-border shopping. A range of excise taxes as well as the general VAT rate were significantly higher in Denmark than in neighboring Germany, implying that many people from especially southern Jutland frequently went to the border to pick up the corresponding goods, especially tobacco, wine, beer, and liquor but in periods also gasoline. This led to recurring downward adjustments in the excise tax rates. At the same time, the 1990s saw the introduction or strengthening of “green taxes,” that is, excise duties on environmental damages or energy use; in fact, at the turn of the century these green taxes contributed significantly to providing the financing of lower marginal tax rates on income.
A tax freeze since the early 2000s. This tax freeze, instituted by the entering right-wing government in 2001, implied a freeze of a host of (rates of) taxes and excises in the system. Both ad valorem tax rates and specific taxes were frozen, in effect causing a gradual dilution of tax intake. The philosophy behind the tax freeze was explained as a desire to enable tax payers to foresee the burden of tax in the future, guaranteeing that there would be no “bad” surprises in the tax area. The full consequences of the tax freeze have unfortunately not been studied in detail. However, some attention has been devoted to a particular aspect of the tax freeze, namely the dilution of the imputed rent tax on housing and its distributional implications. The freeze has caused the imputed rent tax to remain almost constant in nominal terms since 2002, enabling owner-occupiers to save large amounts of money relative to a situation without the freeze (and the continuation of the usual computation of imputed rent taxes prior to 2002). Figure 6 contains average tax savings across income groups for the year 2009. The group of owner-occupiers is divided into ten income deciles. Persons with an income under the median gain only around ½ percent of their incomes, whereas individuals in the top decile gain 3½ percent.
The corporate income tax rate 1987–2011.
Dilution of tax revenue due to the tax freeze.
In addition to the tax adjustments already mentioned in points (1) to (4), many minor adjustments have taken place in the tax system. Of note are perhaps the limitation of deductibility of interest on company debt and the opportunity of applying for group taxation for multinational companies. Further, a series of fringe benefits have become liable to tax.
Scrutiny of the Tax System
In the spring of 2001, the Economic Council in Denmark presented a report featuring an investigation of the Danish tax system and concluding with a long list of proposals for adjustment of the system. The background for the report was twofold: the “internal pressure” on the tax system stemming from population aging and the associated increased need for public expenditure, and the weak projected development of labor supply, and the “international pressure” on the system related to increased mobility of capital, labor, company income, saving, and consumption purchases across borders.
Despite the 2001 analysis and accompanying proposals, not much happened over the next many years, prompting the Council to perform a similar exercise in the latter half of 2008. A bit before, the government had set down a tax commission that delivered its report in early 2009 (see Report from the Danish Tax Commission, 2009). On the basis of these works and extensive political debate, the government launched a “tax reform” in late 2009, taking effect in 2010 and 2011. 17 Most recently, the new center left-wing government that came into office in October 2011 agreed on a tax reform with the two right-wing parties that formed the previous government (more on this reform in the section on The Recent Tax Reform).
Main Problem Areas
The Economic Council’s report from spring 2001 identified a series of problem areas for the Danish tax system. Many of these in essence had to do with an inadequate implementation of the DIT system, in particular the widely different tax treatment of capital income types.
According to the DIT system in its purest form, capital income should be taxed at a common, flat rate. In 2001, personal capital income was, instead, taxed via a complicated progressive schedule. But more importantly, the effective tax on different components of capital income was, and still is, widely different. At the bottom, the effective tax on pension savings was computed to around 20 percent. The effective tax on imputed rent on housing was marginally higher, around 25 percent, and the total effective tax on earnings from share capital in companies again higher (about 40 percent). At the top, the effective tax on interest payments from bank accounts and returns on bonds were at no less than 75 percent for individuals with high net capital incomes. 18 Such a system with wide variation in tax burden on different forms of capital income inevitably causes distortions to people’s placement pattern for their savings. 19,20
In most countries in the EU (as well as in the United States), savings for retirement receive preferential treatment in the tax system. The EU guidelines recommend a so-called EET treatment of savings in pension schemes—exemption of amounts paid in, exemption of return to the stock of savings, and taxation of payouts from the schemes. Denmark actually applies moderate taxation (at 15.3 percent) of the return to pension savings in occupational or individual schemes, so “EtT” would be a better characterization. Despite (symbolic) taxation of returns, preferential treatment is costly for the fiscs, and it is hotly debated to what extent such preferential treatment actually increases total savings in the economy. 21,22
The spring 2001 report from the Economic Council contained, next to its chapter on the Danish tax system, a chapter documenting the distorted and excessively subsidized housing market in Denmark. As interest payments on mortgages were deductible at tax rates ranging between 40 and 60 percent, depending on the individual’s net capital income position, in order to constitute neutral tax treatment of the housing sector imputed rent on housing had to be taxed at a substantial rate. The report estimated that an appropriate rate of tax (measured as a percentage of the value of a dwelling) would be around 2 percent. Instead, most individual’s effective tax on imputed return was well below 1 percent, creating a major tax preference for owner-occupied housing. Since 2001, the tax freeze as mentioned earlier has only aggravated the problem (Economic Council [2011] estimates that the average effective imputed rent tax is 0.5 percent, down from around 0.7 percent in 2001), although the gradual decline in interest rates has worked in the opposite direction. Still, there is no question that owner-occupied housing is a fortiori heavily subsidized.
Finally, the mechanisms of progression in the taxation of personal capital income, of dividends and capital gains from shares, and of imputed rent on housing are entirely uncoordinated.
A disgrace to the tax system in 2001 was a long list of “tax expenditures” (savings in tax relative to a treatment along the normal or general rules in the respective area). The Economic Council reported that such tax expenditures amounted to nearly 3 percent of GDP in 2001. Many of these tax expenditures originally arose as subsidies intended to be transitory and for some reason conceded as preferential tax treatment rather than outright subsidies on the expenditure side of the budget. In any case, these implicit subsidies have stuck and become permanent. The Economic Council recommended a close scrutiny of these tax expenditures, ending in either removal or transfer to the expenditure side as explicit subsidies to be reconsidered on a recurrent basis.
Alongside pointing to the preceding distortions created by differential treatment of incomes in the tax system, the report also contemplated various adjustments in the personal income tax in order to promote labor supply. In 2001, the “population aging” phenomenon had not yet started in Denmark, but it would only be a decade away. Thus, society might as well start to prepare for a situation in which a shrinking population at working age would need to cater for a strongly increased number of elderly people. An obvious solution would be to increase labor supply, either by stimulating entry into the labor force or by stimulating people to work more (full-time rather than part-time, more hours at the margin). For that purpose, different buttons in the tax system—rates, income interval limits, earned-income tax credits—could be considered. Not surprisingly, the biggest effects in terms of raising effective labor supply were associated with raising the income level at which the top marginal tax rate would start to kick in.
While most of the analysis in the Economic Council’s report had little effect on tax policy formation in the succeeding years, the analysis of instruments for increasing effective labor supply did. The right-wing government gradually lowered marginal tax rates for middle and top incomes and introduced an earned income tax credit.
One aspect of the tax freeze mentioned previously in the section on Tax Policy: Reforms and Adjustments since the 1980s was that ad valorem and specific-based taxes were twisted apart as years went by. Opponents would have preferred an “intelligent tax freeze,” in which the relative weight of various taxes would not change this way.
A more worrisome feature of the tax freeze was its impact on the housing market. As detailed previously in the section on Main Problem Areas, mortgage interest payments in Denmark have always been deductible from tax, and at the same time there has been a simple and merely symbolic imputed rent tax. From 2001 onward, the imputed rent tax payments were basically frozen in Danish Kroner, implying that the user cost of owner-occupied homes declined, ceteris paribus. Given that in general the more wealthy families are owner-occupiers, whereas the less well to do are renters, the income inequality effects were questionable, to say the least.
In addition, the tax freeze and the concomitant fall in real user cost put an upward pressure on house prices. 23 At the same time, the financial sector in Denmark produced several new types of favorable mortgage loans (for instance a thirty-year loan with variable interest and with no repayment over the first ten years). These loans were rewarded against expectations of further rapid increases in house prices. Little wonder that a housing price bubble developed, bursting with grave consequences after the onset of the financial crisis. 24
The Recent Tax Reform
After election in mid-September of 2011, a new center-left government came into office in the beginning of October. According to the “package deal” worked out when the government was formed, it aimed to implement a fully financed tax reform that markedly lowers the tax on labor. It also wished to increase the revenue of the state via higher taxes and excises and via restructuring of support of selected industries. At the same time, the government aimed to ensure that the total effect of the tax reform and tax financing had a reasonable social balance.
Since then, the minister of taxation has repeatedly expressed his desire to “pat workers on their shoulders” by lowering their income tax. Accusing the previous government of unfinanced tax reliefs, he has instead declared that tax and excise increases elsewhere in the system will be applied to finance the cuts in marginal income tax rates. Seemingly undecided as to which taxes and excises to raise, he and other members of parties in the government put out several feelers involving an additional tax on people with taxable incomes in excess of one million DKK (US$170,000), strengthening of inheritance taxes, reintroduction of the wealth tax, higher taxation of returns to pension savings, stronger tax burden on housing, including a capital gains tax on housing, and the like. 25
In May 2012, the government issued its plan for tax reform. Thereafter followed an interesting political game; the minority government had to decide whether to make an agreement with its left-wing support party or selected right-wing parties. In the end, and in June, a tax reform agreement was made with two right-wing parties, and the agreement outlined a tax reform to take place in 2013 and subsequent years. 26
The reform aims at reducing taxation of labor income by about 14 billion DKK toward 2022 (when the reform is fully introduced). Major elements of the tax cut are a significant increase in the income level at which the top marginal tax rate kicks in; a doubling of the existing earned income tax credit plus an even higher credit for single parents; and repeal of the so-called entrepreneur tax, the tax on capital gains on shares in private (i.e., not publicly traded) companies. Financing elements are expected cuts in military expenditures and in EU contributions; an increase in the “wage sum tax” on the financial sector (which originally was instituted to counteract the VAT exemption of the sector); indexation of all transfers to price developments rather than wage developments; introduction of price indexation of a series of excise taxes; introduction of means testing of child benefits; limitation of “social dumping”; sharpened taxation of fully taxable individuals’ foreign income; increased taxation of diesel cars and “free cars” (company-provided); and a few others.
Reactions to the reform agreement have been varied and at the same time predictable. The Federation of Danish Industries has expressed its satisfaction with the overall tax burden being reduced. The Federation of Labor Unions has tried to convince the public that many unemployed will actually win from the reform. And the government’s supporting party as well as large groups of members of the left-wing SF and social democratic parties in government have complained that the high earners will gain the most and transfer recipients will come out as big losers. In short, they claim that a reasonable social balance is not attained.
Does the Recent Tax Reform Survive a Trilemma Test?
First, although the parties in the tax agreement have stressed that they aim at a fully financed tax reform, some of the financing elements are fraught with uncertainty. It is not clear that in the end, enough financing will be provided for the drop in tax revenue caused by the tax relief on labor income, including the raised earned income tax credits. A possible line of defense could be that such computations are carried out assuming unchanged behavior, whereas there conceivably are pronounced positive “dynamic effects” of the tax cuts stimulating effective labor supply, which will cause increases in the tax intake more than large enough to cover any financing holes.
Second, and closely connected, is the question whether the reform is going to stimulate growth. Most economists agree that the reform will indeed lead to an increase in effective labor supply, as more people are led into the labor force, and those who can, may wish to work more hours. The extent to which labor supply may be boosted is, however, quite uncertain; there is regrettably too little firm knowledge about behavioral changes at the external and internal margins.
In a recent working paper on the responsiveness of taxable income responses to Danish tax reforms, Kleven and Schultz (2011) conclude that labor income elasticities are modest overall, around 0.05 for wage earners and 0.10 for self-employed individuals, and that behavioral elasticities are much larger when estimated from large tax reform episodes than for small tax reform episodes. They further claim that their results are quite robust. Using their results one should probably only expect a modest increase in taxable income following the income tax reliefs (which for most people involved would be rather small). Nevertheless, the government expects an increase in labor supply corresponding to about 15,000 full-time employed persons, or some two-thirds of a percentage of the labor force. 27
Aside from its effects on the labor force, will the reform be able to stimulate entrepreneurship and innovative activity in established companies? Not much in the reform is working in that direction. Moreover, the economic literature on the relation between taxation and innovation and entrepreneurship is not entirely clear. 28 A further complication is that as stressed in the section on Economic Growth and Productivity in Denmark, the recent Danish productivity development is not well understood. It is conceivable, though, that the removal of the so-called entrepreneur tax (the tax on capital gains on closely held shares) can foster some innovation and growth, even though it is not very precisely targeted. For theoretical arguments see Keuschnigg and Nielsen (2004a, 2004b).
Third, does the reform score well in terms of promoting a more equal distribution of income and consumption opportunities in the population? No. The change in the indexation rule for transfer payments (unemployment benefits, welfare payments, public pensions, etc.) away from following wage developments in the public (and private) sector to following inflation will over time bring about a major relative drop in consumption possibilities of those individuals who are most dependent on transfers. At the same time, people with high incomes will enjoy marked increases in disposable incomes on account of the shift in the top marginal tax rate interval. And since remaining initiatives have less effect on distribution, overall the reform just cannot improve equality in society, on the contrary.
Newly issued calculations from the Ministry of Finance 29 support the conclusion. The calculations report the effect on disposable incomes of twenty-six different family types in 2023, when the reform is fully phased in. Considerable losses will be registered among people on welfare payments, unemployment benefits, and early retirement pensions, while the biggest gains will accrue to privately employed and highly educated persons. 30
In sum, then, while the reform may score a few points in terms of contributing to raising growth, and while including dynamic effects may imply that it is not underfinanced after all, it cannot pass the distributional branch of the trilemma test, so must fail the test overall.
Sensible Reform Elements and the Trilemma Test
In a working paper from the Danish Ministry of Finance, Jensen (2001) wrote that “Possible future tax reforms will involve significantly harder trade-offs between efficiency and distributional objectives, as the scope for broadening the tax base further is limited.” In other words, he forecasted that tax reforms after 2000 would have a hard time passing the trilemma test.
As mentioned earlier, in a couple of rounds in the 2000s tax changes were carried out in which marginal tax rates have been cut and earned income tax credits introduced and extended. The distributional consequences of these tax changes have not been studied in full.
The recent tax reform, laid out in the section on The Recent Tax Reform, likewise confirms Jensen’s prediction. It seems difficult to single out reforms without very hard trade-offs between efficiency and distributional objectives. Nevertheless, we shall try to point to some elements of tax changes that obviate the dilemma.
Look back to the Economic Council’s report from 2001, cfr. Main Problem Areas. Major problems in the tax system laid out in that report were the uneven tax treatment of different forms of capital income, the lenient taxation of pension savings, the heavy subsidization of the housing sector, and the many tax expenditures. Very little has been done in these areas since then, rendering them obvious candidates for components of tax reforms that may bring about efficiency gains without hurting distribution.
The first branch of the trilemma test concerns sustainability of fiscal policy. Presuming that no tax is on the wrong side of the Laffer curve, every cut in a tax must be met by an increase in some other tax (or an expenditure cut, for that matter).
As to the second branch of the trilemma test, both OECD (2008) and Prammer (2011) end up with a ranking of different taxes according to how much they distort and thereby hamper growth. Least distorting are property taxes, including taxes on housing; then come consumption taxes; then personal income taxes; and finally corporate income taxes. Thus, one possible guide for tax reforms would entail putting more weight on property and similar taxes and less weight on personal and company income taxes.
When it comes to the distributional branch of the trilemma test, we have seen how implicit subsidization of owner-occupied housing strongly benefits the well-off individuals. Moreover, one can imagine an untapped potential in other wealth type taxes with beneficial distributional characteristics.
Against this background, a reform with overall favorable trilemma features might contain, on the financing side, (1) strengthening of the imputed rent tax on housing. Do away with the tax freeze effects and gradually increase the tax rate toward a level implying neutral tax treatment of owner-occupied housing. (2) Removal of a major part of tax expenditures and transformation of those that still have a sensible role to explicit subsidies on the expenditure side. At the moment, tax expenditures still amount to around 40 billion DKK (about 6½ billion USD) or approximately 2.2 percent of GDP; see Ministry of Taxation (2010) and Terkilsen et al. (2012). (3) Perhaps increase in the taxation of land, for instance, in the form of a state tax on land. (4) Stronger taxation of industrial foundations; currently, these are taxed like corporations but can effectively deduct donations at a 125 percent rate from their taxable incomes. (5) Perhaps strengthening of estate and inheritance taxes. (6) Attempt to put debt and equity on par in the corporate income tax system. Deductibility of interest on debt, but no similar deductibility of normal return to equity creates complications and distortions. Either go for a comprehensive business income tax type (as once proposed by the US Treasury) or an allowance for corporate equity type (as proposed by the Institute for Fiscal Studies, London, and in the Mirrlees et al. [2010] review) corporate income tax. (7) Moves toward more equal treatment of different capital income components. Depending on revenue contributions from the preceding measures, an overall lowering of capital income tax rates might be in order.
On the outlays side, (1) do away with the price regulation of public transfer incomes and return to wage regulation. A once-and-for-all adjustment of transfer income levels is not ruled out, but regulating transfers according to price and not wage developments is bound to create tensions in society. (2) Conduct smart adjustments in the personal income tax. Smart meaning here efficiency-enhancing; they do not have to score points on distribution, as so many other elements in a reform can take over.
Conclusion
This article has taken a view at Danish tax policy, recent tax reforms, and outstanding problems in the tax area through the trilemma lens; that is, it has looked at the fiscal sustainability, growth and innovation, and distributional dimensions of taxation.
It claims that, after the introduction of a version of the DIT system with its many favorable characteristics, subsequent tax changes and reforms have not all scored well in the trilemma test. The most recent reform in 2012, for example, preserves sustainability of public finances and may give a small push to labor supply, innovation, and growth. But it fails the distributional part of the test.
The article acknowledges that it may have become increasingly difficult to design tax changes in Denmark that stands a fair chance of surviving trilemma scrutiny. But it argues that there are still a number of components that, upon introduction into a tax package, should be able to help out. Both the current government and its immediate predecessor have been in “reform mode,” adjusting in turn unemployment benefits, early retirement provisions, welfare payments, and student grants. So perhaps the tax system will be given yet another look.
Footnotes
Acknowledgments
We thank the editors James Alm and Steven Sheffrin as well as participants in the Fiscal Trilemma conference for comments and suggestions.
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.