IB revija 1-2/2006       UMAR    111

	









  	 	      
%!# *!!9))
;

!



8*!.<!= 
+*!.<!=
Thomas Larsen*
A Flat Tax in Denmark?
1. Introduction
The Danish tax system is a classic system with
progressive taxation of labour income and positive
net capital income. Compared to most countries
the Danish system is characterised by relatively
high marginal tax rates at relatively low-income
brackets and almost no social security contributions
as a means of financing public expenditures.
Furthermore the Danish tax base is to a wide
extent based on comprehensive income with the
major qualification that most tax credits are not
(fully) deductible in the base of the state and
progressive income taxes. This is especially true
for net negative capital income (i.e. interest
payments), which is treated more in line with a
dual income tax system.
High marginal taxes and complicated rules for the
calculation of taxable income has led to public
debate on the possibilities and effects of introducing
a flat tax in Denmark. So far the debate has stranded
due to the potentially very big distributional effects
of such a tax reform. Also the debate in Denmark
has not been very clear as to whether one was
referring to a flat tax rate or a flat tax.
This paper starts by describing the existing Danish
tax system in sections two to four. Some parts of
the system might seem to be very far from a f lat
tax while other parts comes close to the archetypical
flat tax. In fact – measured in revenue – the Danish
tax system is not very far from being a flat rate tax
system. But the distributional effects from a revenue
neutral transmission from the current tax system
to a flat tax system seem to be quite large compared
to what historically has been accepted when
reforming the Danish tax system.
In section five I show the effects of three different
revenue neutral f lat tax experiments to illustrate
the difficulties in implementing a flat tax reform.
The experiments shown in this paper are based on
calculations made on the Danish micro simulation
model. There are no estimates of effects on labour
supply etc.
Finally in section six I briefly discuss the trends
in the development of the Danish tax system and
take a look at what the future might bring.
2. Personal Income Taxation
2.1. Labour Income
The Danish taxation of labour income is
progressive following the schedule shown in figure
1. Total marginal tax rates vary from 8 % to 63 %.
An 8 % labour market contribution is levied on
all labour income in practice with no tax credits.
Fig. 1.: Marginal and average tax rates, 2005.
* Ministry of Taxation, Denmark (Head of Division, Economic Analysis
112     UMAR      IB revija  1-2/2006
On top of the labour market contribution muni-
cipal and county taxes of 33.3 % on average are
levied on all taxable income above a personal
allowance of DKK 37,600 (  5,035). Taxable
income includes both labour, most transfers and
some capital income with credits for interest
payments, expenses aiming at maintaining income,
membership of labour union etc. Also an in work
tax credit is deducted as well as the labour market
contribution.
The basic state tax of 5.48 % is levied on all
personal income including positive net capital
income with credits for the labour market
contribution, pension savings and expenses for self-
employed.
Fig. 2.: Number of average and marginal taxpayers
	
		

			

		
			
		

			
	 


 	
 
	

 


 	
 
 	
 	

 
The middle (6 %) and top (15 %) state taxes are
levied on the same income as the basic state tax
above thresholds of DKK 259,500 (   34,740) and
DKK 311,500 (   41,700) respectively. It is allowed
to transfer unexploited personal allowance and
middle state tax threshold to the spouse. The
average taxes paid are shown in table 1. In figure 2
the resulting average and marginal tax rates are
illustrated.
Part of the argument put forward by some in favour
of a flat tax in Denmark is that it is difficult to
calculate taxable income and tax due to the
differences in tax base, tax credits that may be
deducted in the different tax bases and the different
tax rates applicable to the different tax bases.

(
/
)
+!
./) 9

!) ())
=
 (
/
 =
 (
/
 =
 (
/

9)) 1	52 9)) 1	52 9)) 1	52 9))
,%#!	5 	
 	5	  5  	5

	5	B	5  5 
 5

 
 5
	5	B5 		 5	 
 5  5
5	B5 
 5  

5  5
5	B5  5	 
 
5  5
5	B5 	 		
5
 
 	5

  		5

5	B5 	 	
5 
 	5
 	
 	
5
5	B5 
 	

5
 	
 	5
 
 	

5
5	B
5 
 5 	 	5  5

5	B
5 	
 5 	
 
5
  	5

5!B 
 
5 			 
5 		 	5	


 5	 	5 	5 	5 5	
 		5
€
€
€
IB revija 1-2/2006       UMAR    113
At a first glance this might seem to be a convincing
argument. In reality however the vast majority
of Danish tax payers spend little or no time doing
their tax returns. Tax authorities receive informa-
tion on income and tax credits from employers
and banks etc. and do the calculation for the
taxpayer. Thus the actual gains in reduced
administrative costs to tax payers would be
relatively modest compared to the current tax
system.
The revenue from the different parts of the personal
income taxation is shown in table 2.
The total Danish tax burden is around DKK
765,000 million (  102,410 million) and GDP is
around DKK 1,540,000 million (  206,160 million)
in 2005 leaving the tax-to-GDP ratio at 49.6 %.
Personal income taxes amount to 23.5 % of GDP
and 47 % of total taxes.
2.2. Capital Income
The Danish capital income tax system is quite
complicated and tax rates on net positive capital
income are very high. In some cases real tax rates
exceed 100 percent. The lowest tax rates vary from
0 % on gains on owner occupied housing and 15 %
on the yield from pensions savings. In the middle
range taxes on shares are taxed progressively with
either 28 or 43 %. Finally interest payments from
bank accounts and bonds are taxed with the
progressive income taxes mentioned above with a
maximum of 59 %.
Net negative capital income is deductible in the
municipal and county tax with an average rate of
33.3 %. Since the 1980’ies this rate has been
reduced substantially from 73 %.
:2elbaT lanosrepfoytirojamehtmorfeuneveR
noillim000,1KKD,5002nisexatemocni
:xatemocnietatS 0.56
xatcisaB- 2.24
xatelddiM- 6.7
xatpoT- 2.51
:noitubirtnoctekramruobaL 2.76
:sexatytnuocdnalapicinuM 1.032
sexatlapicinuM- 7.541
sexatytnuoC- 8.86
renwonoxatytnuocdnalapicinuM-
sesuohdeipucco 8.01
xathcruhC- 8.4
sexatemocnilanosrepmorfeuneverlatoT 3.263
3. Corporate Taxation
The Danish corporate tax rate is 28 %. The rate
has been reduced continuously from 50 percent
to the current level since the 1980’ies. Rate
reductions have been followed by base broadening.
Revenue is currently around DKK 50,000 million
(   6,695 million) corresponding to 3.1 % of GDP
or 6.3 % of total taxes. In addition there are specific
taxes on the extraction of hydrocarbons.
4. VAT and Excise Duties
The Danish VAT-system only operates with one rate
of 25 % on almost all goods and services. There are
no reduced rates although some services are exempt
or zero-rated in accordance with EU regulation.
Revenue is currently around DKK 155,000 million
(    20,750 million) corresponding to 10 % of GDP
or 20 % of total taxes.
The Danish Tax system also includes a wide range
of excise duties on cars, energy, tobacco, alcoholic
beverages, waste and water etc. These excises are
used to pursue several different goals being both
fiscal as well as goals within different policy areas
like environment, traffic, energy and health.
Revenue is currently around DKK 95,000 million
(  12,715 million) corresponding to 6 % of GDP
or 12 % of total taxes.
5. Modelling Flat Tax in Denmark
It is possible to model at least three different main
categories of flat tax experiments in the Danish
economy that have been touched upon in the
public debate. Other variations of calculation of
income, size of basic allowance and tax rates could
have been considered. The effects of the following
three revenue neutral f lat tax experiments are
analysed:
A. A flat tax of 35.2 % on all positive income.
The only tax credit is for pension savings.
B. A f lat tax of 42.0 % on all positive income
above a basic allowance of DKK 37,600. The
only tax credit is for pension savings.
C. A flat rate of 48.9 % on existing taxable income
above a basic allowance of DKK 37,600.
These experiments might seem simple. In reality
however, a f lat tax reform might be extremely
€
€
€
€
€
114     UMAR      IB revija  1-2/2006
-30
-20
-10
0
10
20
30
40
50
60
70
B
elow
100,000 
100,001-
150,000
150,001-
200,000 
200,001-
250,000 
250,001-
300,000 
300,001-
350,000 
350,001-
400,000 
400,001-
500,000 
500,001-
600,000 
600,001-
800,000
800,000 -  
Household income, DKK
Pe
r 
ce
nt
Flat tax 35.2 per cent Flat tax 42.0 per cent Flat tax 48.9
complicated to implement, as most income
transfers are gross of taxes and would have to be
recalculated to take into account the changes in
tax rates. Also the treatment of self-employed and
majority shareholders should be addressed. And
it would be necessary to decide upon the treatment
of existing pension savings.
In figure 3 the resulting average tax rates are shown
with increasing income. As is the case in figure 1,
the shown tax rates are not taking into account
possible tax credits. Results from micro simula-
tions on a sample of the actual Danish population
of taxpayers are shown in table 3-5. The changes
in average taxes paid are shown in figure 4.
In the first experiment (A) the marginal and
average tax rates are equal. This flat tax experiment
will shift more than DKK 30,000 million ( 4,015
million) in tax revenue from poorer households
to more wealthy households. More than 2.3
million Danish households will loose an average
of DKK 12,735 ( 1,705) a year. Only 780,000
households will gain on average DKK 39,430
( 5 ,275) a year. In general households with income
of less than DKK 350,000 (  46,855) will loose.
Households with income of more than DKK
350,000 will gain.
The loss (gain) will be reduced (increased) as
income increases. Thus the result of the experiment
is highly regressive. The poorest households face
an increase in average taxes paid of 68 % while at
the other end of the income scale the wealthiest
households receive average tax cuts of 25 %.
In the second experiment (B) the marginal tax rate
is 42.0 % above a personal allowance of DKK
37,600. This of course introduces progression in
the tax schedule and average tax rates increase
0
5
10
15
20
25
30
35
40
45
50
0 40.000 90.000 140.000 190.000 240.000 290.000 340.000 390.000 440.000
Wages, DKK
Per cent
Flat tax of 35.2 pct. Flat tax of 42.0 per cent Flat tax of 48.9 per cent
Fig. 3.: Average tax rates of three flat tax experiments.
Fig. 4.: Changes in average taxes paid.
€
€
€
€
IB revija 1-2/2006       UMAR    115
gradually from 0 towards 42.0 %. The effect of
this change compared to experiment (A) is that
only around DKK 20,000 million (  2,675
million) shifts from lower and middle-income
households to high-income households. On average
more than 2.2 million households loose DKK
8,845 (  1,185) while 820,000 households gain
DKK 24,980 (  3,345). Because of the personal
allowance very low-income married couples
actually gain. The introduction of a personal
allowance does repair some of the distributional
problems faced in experiment (A). Both losses and
gains from the experiment are reduced. The losses
are no longer decreasing with increasing income
due to the introduced progression in the tax schedule.
The downside of this perhaps less unacceptable
distribution of taxes is that only households with
income of more than DKK 600,000 (  80,300)
will gain. The lower to middle income households
face tax increases of around 10 % of current average
taxes while the high income households receive
tax cuts of 3-17 % on average.
The third experiment (C) is closer to the current
Danish tax system in the sense that the tax base is
€
€
€
€

,

!/)
)

(#
+!
./) 9

!) ())
=
 (
/
 =
 (
/
 =
 (
/

9)) 1	52 9)) 1	52 9)) 1	52 9))
,%#!	5 	
 
5  B5		  
5

	5	B	5  	5	 
 5
 
 	5

	5	B5 		 5
 
 5  	5
5	B5 
 
5

  	
5
  
5
5	B5  5 
 5  5

5	B5 	 B	5
 
 5
  5
5	B5 	 B		5 
 	
5
 	
 B	
5	B5 
 B
5
 	
 		5
	 
 B5
5	B
5 
 B5
 	 
5  B	

5	B
5 	
 B
5 	
 B	
5	  B5	


5!B 
 B5 			 B		5
 		 B	5



 5	 5
 	5 B
5 5	
 B

3

!/)
)

@#
+!
./) 9

!) ())
=
 (
/
 =
 (
/
 =
 (
/

9)) 1	52 9)) 1	52 9)) 1	52 9))
,%#!	5 	
 B  B	5  B

	5	B	5  5	 
  
 5	
	5	B5 		 5 
 5	  5	
5	B5 
 
5	  5  5

5	B5  
5
 
 5	  5	

5	B5 	 5 
 		5	  
5	
5	B5 	 B5 
 	5 	
 5	
5	B5 
 B	5	 	
 	5 
 	5
5	B
5 
 B
5 	 	5		  
5


5	B
5 	
 B
5 	
 B5
  B5

5!B 
 B	5
 			 B
	5

 		 B

5


 5	 	5	
 	5 B5
 5	
 B	
116     UMAR      IB revija  1-2/2006
the familiar tax base used for municipal and county
taxes including deductions for interest payments
and other tax credits etc. The marginal tax rate is
48.9 % above a personal allowance of DKK 37,600.
Again this experiment has a progressive tax
schedule. The effect of this experiment is to shift
more than DKK 18,000 million (  2,410 million)
from lower and middle-income households to high-
income households. More than 2.3 million
households loose an average of DKK 8,13 0  (
1,090) while 762,000 households gain an average
of DKK 23,945 (  3,205). Again, because of the
personal allowance, very low-income married
couples actually gain.
Holding on to the tax credits in the current tax
system means that the tax rate has to be increased
compared to the other f lat tax experiments. For
household income below DKK 250,000 (
33,465) the effect of the higher tax rate dominates
the effect of the tax credits. These households face
increasing losses compared to experiment (B), but
still the losses are lower than experiment (A).
Households with incomes between DKK 250,000
and DKK 800,000 (  107,095) are better of than
in experiment (B). This is because the average and
marginal tax rate of 48.9 % is relatively close to
the tax rate they face in the current tax system and
in fact lower for the more wealthy households.
Keeping the tax credits dominates the effect of the
relatively high tax rate. Finally the households with
the highest incomes will gain. But less than in
experiment (A) and (B). The reason for this is
again that the higher tax rate dominates the effect
of keeping the tax credits.
The low-income households face average tax
increases of 10-20 % while the higher income
households receive tax cuts of 1-15 %.
The distributional effects shown above are of course
calculated before any possible behavioural effects
on labour supply, education etc. I have not tried
to assess these effects, but they would have to be
considerable in order to sufficiently reverse the
first order distributional effects from the tax
experiments. It is however possible to make one
or two notions about what could be expected from
effects on labour supply and education.
In the case of experiment (A) almost every
household and taxpayer faces lower marginal tax
rates. This indicates that labour supply will increase
substantially from this kind of tax reform. Also
reduced progression will work in favour of
increases in the level of education. The costs of
these effects are that more than 3 million taxpayers
face increased average taxes.
Experiments (B) and (C) try to repair some the
deficiencies in experiment (A). In experiment (B)
around 2.8 million taxpayers out of 4.46 million
taxpayers face reduced marginal tax rates while the
rest of the taxpayers face mostly smaller increases.
In experiment (C) only around 1 million taxpayers
face reduced marginal tax rates while around 3
million taxpayers face larger increases in the
marginal tax rates. While these tax experiments most
probably will have positive effects on the incentives
to increase education the effects on labour supply
is less favourable than experiment (A).

B

!/)
)

&#
+!
./) 9

!) ())
=
 (
/
 =
 (
/
 =
 (
/

9)) 1	52 9)) 1	52 9)) 1	52 9))
,%#!	5 	
   B	5  B	
	5	B	5  
5
 
 
 
 
5
	5	B5 		 5
 
 
5  
5
5	B5 
 
5  
5
  
5	
5	B5  5	 
 	5  
5


5	B5 	 !5	 
 	5	  5

5	B5 	 B5 
 	5 	
 5
5	B5 
 B	5	 	
 5 
 B	5
5	B
5 
 B5		 	 5  B

5	B
5 	
 B	5
 	
 B
5	  B		5

5!B 
 B	
5
 			 B	5
 		 B5


 5	 5 	5 B5
 5	
 
€
€
€
€
€
IB revija 1-2/2006       UMAR    117
6. Is There a Future For the Flat Tax in
Denmark?
This paper has shown the effects of three of the
most straightforward revenue neutral f lat tax
experiments that could be envisaged in Denmark.
The conclusion to this analysis is that Denmark is
not likely to introduce a f lat tax due to the
distributional effects from such a tax reform.  If
you – like Denmark – at the outset have a tax system
with relatively high levels of progression it will be
very difficult to shift towards a flat rate tax system
without reducing tax revenue. And the total tax
reductions required, might have to be substantial.
However, it is very likely that a flatter tax system
will gradually emerge during the coming years.
Most parties in the Danish Parliament have plans
to reduce marginal and average taxes on labour
income. The controversies are mainly related to the
financing of the tax cuts and to the mix of the tax
cuts between high, low and middle-income earners.
The current government has as a goal to reduce
taxes on labour income when it is compatible with
sustainable economic conditions. The Danish
Welfare Commission has recommended lower
taxes on labour income and positive capital income.
The development since the 1980’ies has shown a
history of continuous cuts in marginal labour tax
rates. Globalisation and the aging population put
increasing pressure on the tax system to promote
labour supply, education, R&D etc.
Finally, there are already elements in the current
Danish tax system that could be thought of as
being similar to a flat tax. For instance the uniform
VAT-rate and large parts of the personal income
taxation.
Measured as the share of total tax revenue capital
taxation is not far from being in effect a flat tax.
Reduced rates on parts of positive capital income
at gross costs of 1 % of tax revenue would be
sufficient to change the taxation of capital (except
for housing and pension funds) into a flat tax
system. A similar cut in progressive taxes on
labour income removing middle and top state
income taxes would require gross revenue cuts of
around 3 %.