1

Mmmm

Introduction

Modern states could not exist without tax systems which raise large amounts
of revenue to pay for public services. Most take in excess of 30% of national
income in tax. Some take nearly half. The way in which these systems are
designed matters enormously to economic welfare. Yet policymakers rarely
step back to consider the design of their national tax systems as a whole.
Public understanding of taxation is limited. And the political and public
discussion of tax design is woefully inadequate.
Tax by Design is both an imperative and a description of our approach in
this review. Our aim is to set out the principles on which a 21st century tax
system should be based and then to apply them in suggesting concrete policy
recommendations to improve the UK tax system. To that end, we use
insights from economic theory and empirical research to discuss the impact
that the tax system has on people’s behaviour, and the resulting trade-offs
that policymakers have to make between the various and often conflicting
objectives that they might wish the tax system to achieve.
In doing so, we follow in the footsteps of the Nobel laureate James Meade,
who chaired a previous review of the tax system for the Institute for Fiscal
Studies three decades ago. The Meade Report1 focused on the structure and
reform of direct taxation and has been influential in debates over tax policy
ever since. Dauntingly, our canvas is wider than that, covering the whole tax
system and some areas of interaction with the social security and tax credit
systems.

1
Meade, 1978.
2 Tax by Design

Recognizing the importance of law, accountancy, politics, psychology, and

other approaches, we nevertheless approach these issues through the prism
of economics. This provides us with a framework that allows us to ask, for
example: What are we trying to achieve? How does the structure of the tax
system influence people’s behaviour? What are the economic costs and
benefits of particular policy choices? And who gains and loses from them?
The tax system is, of course, both enormous and fearsomely complex. Tax
legislation in the UK runs to over 8,000 pages, and the books that lawyers
and accountants use to interpret it run to millions of words. Confronted with
that size and complexity, we try to be as detailed in our analysis and
recommendations as is necessary to guide real policy decisions, without
getting tangled in the undergrowth that is comprehensible only to specialists.
This means that there will always be further avenues to explore and difficult
special cases to consider. But these difficult cases should not drive the overall
design of the system, even if they need to be accommodated by it.
The primary task we have set ourselves is to identify reforms that would
make the tax system more efficient, while raising roughly the same amount
of revenue as the current system and while redistributing resources to those
with high needs or low incomes to roughly the same degree. Our motivation
is not to achieve textbook tidiness for its own sake, but to unlock significant
potential welfare gains. To the extent that these gains show up as higher
national income, they would also allow the government that achieves them
to loosen the constraints under which the tax system operates—in other
words, to spend more on public services or to redistribute more without
lowering post-tax incomes in aggregate.
Reforming the tax system may not be easy or popular in the short term,
but it holds out the prospect of significant economic gains and hence the
promise of higher living standards in the long term.
Our conclusions on reform are guided by three key considerations. First is
the importance of taking account of the actual economy and population on
which the tax system operates. Taxes apply to people and businesses in the
world as it is, not as we might wish it to be. A tax system that might have
been ideal in the middle of the 20th century will not be ideal for the second
decade of the 21st century. Second is the crucial insight that the tax system
needs to be seen as just that—a system. While we will often address the
impact of each tax separately for simplicity of exposition, we focus
 Introduction 3

throughout on the impact of the system as a whole—how taxes fit together

and how the system as a whole achieves government’s goals. Third, we base
our analysis on the modern economics of taxation. This allows us to develop
a systematic conceptual approach that joins together our thinking across the
whole range of taxes. What we do is rooted in economic theory that models
the constraints people face and the way they behave when taxes change. Our
approach is also determinedly empirical, drawing upon the best available
evidence on the effects taxes have in practice.
While we have a very broad canvas, there are some important issues on
which we deliberately do not take a stance. For example, we do not
recommend what the overall level of taxation should be. The economic
issues involved in this decision are huge, and in many ways fundamentally
different from those involved in designing a tax system. The choice also
involves political judgements about the appropriate role and scope of the
state. Similarly, we do not take a view on the extent to which the state should
seek to redistribute income and wealth from rich to poor. That again is a
primarily political choice, although it does of course have economic
consequences. But we do try to suggest how the state might best use the tax
system to raise more or less revenue—or to redistribute more or less income
and wealth—if the government of the day wished to do so.
We go about this by looking, chapter by chapter, at how to tax earnings,
spending, savings, wealth, housing, and companies. In the next chapter,
though, we consider the economic approach to tax reform and, specifically,
some of the issues in designing the tax system as a whole. And we conclude
by putting forward a long-term strategy and package of reform in Chapter
20.
In this introductory chapter, we provide just a little context which it is
important to understand before we get on to the economic arguments and
the analysis of each type of tax. First, there is a very high-level overview of
the UK tax system and how it has evolved. Second is a quick look at some of
the key changes to the economic environment in which the tax system has to
operate. And third, because tax policy is made in a deeply political
environment, we briefly address some of the political context and constraints
on policymaking.
4 Tax by Design

1.1. THE EVOLUTION AND STRUCTURE OF THE UK TAX

SYSTEM

Many features of the UK tax system today would be familiar to a visitor from
the late 1970s. The government still raises the bulk of its revenue from taxes
on income, spending, and corporate profits and from local property taxes. At
this level of generality, there are important similarities across most industrial
countries.
Nevertheless, there have been some dramatic changes. Value added tax
(VAT) has gained in importance relative to excise duties, the main rate
having more than doubled. The income tax system has moved from joint
assessment of married couples to independent assessment of individuals.
Income tax rates have come down dramatically. The infamous top levels of
83% on earned income and 98% on unearned income have fallen to 50%—
and in fact stood at 40% for most of the period. The basic rate of income tax
has fallen from 33% to 20%. Rates of National Insurance (NI) contributions
have risen, the ceiling for employer contributions has been abolished, and
the structure of NI has become more closely aligned with the income tax
system. Tax credits have been introduced and expanded on a grand scale.
Taxation of savings has been reformed and somewhat improved. The
structure of corporation tax has been overhauled: the main rate has been cut
from 52% to 26% in 2011 with further cuts to 23% due by 2014; tax credits
have been introduced for research and development (R&D) spending.
Capital transfer tax has been replaced by inheritance tax. Domestic rates
have been replaced by council tax, via the brief and disastrous experiment
with the poll tax (or community charge).
Table 1.1 summarizes UK revenue forecasts for 2011–12. Nearly two-
thirds of all tax receipts come from just three taxes—income tax, NI
contributions and VAT. Corporation tax accounts for nearly another 9%.
Fuel duties, council tax, and business rates raise just under another 5%

Notes and Source for Table 1.1:

a
Most of the cost of tax credits is counted as government spending rather than a reduction in income tax
revenue and so is not included in this table.
b
Consists of Carbon Reduction Commitment, social tariffs, feed-in tariffs, and Renewables Obligation.
Note: Figures may not sum exactly to totals because of rounding.
Source: Office for Budget Responsibility, 2011, table 4.7 and supplementary tables 2.1 and 2.7 (available at
http://budgetresponsibility.independent.gov.uk/category/topics/economic-forecasts/).
 Introduction 5

Table 1.1. Sources of UK tax revenue, 2011–12 forecasts

Source of revenue Revenue (£bn) Percentage of taxes (%)

Income tax (gross of tax credits) 157.6 28.0

a
Tax credits counted as negative income tax –4.7 –0.8
National Insurance contributions 100.7 17.9
Value added tax 100.3 17.8
Other indirect taxes
Fuel duties 26.9 4.8
Tobacco duties 9.3 1.7
Alcohol duties 9.7 1.7
Vehicle excise duties 5.9 1.0
Air passenger duty 2.5 0.4
Insurance premium tax 2.9 0.5
Betting and gaming taxes 1.6 0.3
Climate change levy 0.7 0.1
Landfill tax 1.2 0.2
Aggregates levy 0.3 0.1
Environmental leviesb 1.8 0.3
Customs duties 3.3 0.6
Capital taxes
Capital gains tax 3.4 0.6
Inheritance tax 2.7 0.5
Stamp duty land tax 5.8 1.0
Stamp duty on shares 3.3 0.6
Company taxes
Corporation tax (net of tax credits) 48.1 8.6
Petroleum revenue tax 2.0 0.4
Business rates 25.5 4.5
Bank levy 1.9 0.3
Council tax (net of Council Tax Benefit) 26.1 4.6
Licence fee receipts 3.1 0.6
VAT refunds to public sector bodies 15.0 2.7

Other taxes 5.6 1.0

National Accounts taxes 562.4 100.0

6 Tax by Design

apiece, with a range of other taxes accounting for the remainder. These
include indirect taxes on cars, alcohol, tobacco, betting, and various
polluting activities, which between them raise around 6% of government
revenues.
These shares have not changed a great deal over time. For example, at just
under 30%, income tax raises much the same proportion of total tax revenue
now as in the late 1970s, despite the cuts in rates. The most significant
changes have been in the composition of indirect taxes, with VAT raising a
larger proportion and excise duties a smaller proportion of total revenue.2
Similar trends can be observed internationally. Between 1975 and 2008, the
proportion of OECD tax revenues coming from ‘general’ consumption taxes
rose from 13% to 20% while the proportion coming from ‘specific’
consumption taxes fell from 18% to 10%.3
By international standards, the UK raises more than most countries from
income taxes—30% against an OECD average of 25%—and less than average
from social security contributions—19% in 2008 against an OECD average of
25%.4 One feature of the UK system that is unusual by international
standards is its degree of centralization. In the UK, only council tax, which
accounts for less than 5% of total revenue, is collected locally. Only Ireland
has a smaller proportion of taxes administered below the national level.
Local government in the UK is, therefore, funded to an unusually large
degree by central government grants, with local taxes playing a
comparatively limited role.
So, at the macro level, tax systems have many similarities in terms of how
they evolve over time and between countries. But structures differ a great
deal, especially in their details. It is with these structures and design features
that we are concerned—what is the measure of income on which income or
corporate taxes are levied; how are savings treated; how do personal and
corporate tax systems fit together; how progressive is the system; what is the
base for indirect taxes and how are they designed?
By getting these design features right, all countries can reap very valuable
dividends in terms of both increased economic efficiency and greater

2
Details of changes over time can be found at http://www.ifs.org.uk/fiscalFacts/taxTables.
3
OECD tax revenue statistics, table C, http://www.oecd.org/document/60/0,3746,en_2649_
34533_1942460_1_1_1_1,00.html#A_RevenueStatistics.
4
OECD comparative tables, http://stats.oecd.org/Index.aspx?DataSetCode=REV.
 Introduction 7

fairness. And, to be fair, within the UK there has been progress towards a
better system over the 30 years since the Meade Report was published. Tax
administration has improved with the use of technology. We no longer have
wholly ineffective tax rates of 98% on unearned income. The taxation of
savings has been much improved. Taxation of owner-occupied housing has
been rationalized. National Insurance contributions and corporation tax are
now more broadly based. And, for all their unnecessary administrative
problems, the introduction and extension of tax credits have helped improve
work incentives, at least for some groups.
By international standards, the UK system has relatively few loopholes and
opportunities for avoidance. For most people, for most of the time, the tax
system works: it is not overly intrusive and it does not require vast effort to
comply with—although some people on tax credits, in self-employment, or
with complex financial affairs may disagree. We would certainly not
characterize the British system as brutally as some characterized elements of
its US counterpart back in 1995: ‘The federal income tax is a complete mess.
It’s not efficient. It’s not fair. It’s not simple. It’s not comprehensible. It
fosters tax avoidance and cheating. It costs billions of dollars to administer.
… It can’t find ten serious economists to defend it. It is not worth saving.’5
But the UK system is still unnecessarily complex and distorting. Tax policy
has for a long time been driven more by short-term expedience than by any
long-term strategy. Policymakers seem continually to underestimate the
extent to which individuals and companies will respond to the financial
opportunities presented to them by the tax system. They seem unable to
comprehend the importance of dealing with the system as a whole. And real
and effective reform remains politically extremely difficult.
The litany of poor (and expensive) tax policy decisions is a long one. It
includes successive changes to the structure of capital gains tax, the
introduction and abolition of a 10p starting rate of income tax, the
introduction and abolition of a 0% rate of corporation tax for small
companies, tax breaks for film-making (which were estimated by HM
Treasury to have cost an astonishing £480 million in 2006–07), and the
introduction and abolition of the ‘poll tax’. Moreover, the failure of political

5
Hall and Rabushka, 1995, 2.
8 Tax by Design

will means that council tax bills in England and Scotland in 2011 still depend
on estimates of the relative values of different properties in 1991.
These are the issues on which the rest of this book concentrates. But before
focusing down on tax design, it is important also to raise our sights towards
the economic context in which tax design must occur.

1.2. THE CHANGING ECONOMIC CONTEXT

Tax systems need to be designed for the economies in which they are to
operate. Developing economies often need to put a very heavy weight on
collectability of taxes. Economies rich in natural resources need a tax system
designed to reflect that. Highly federalized countries will have tax systems in
which the setting of taxes at the sub-national level is a major concern. In the
UK context, two changes have been so profound over the past three decades
that they really do deserve some special attention. The first is the great
increase in inequality and associated changes to the labour market. The
second is the change in the structure of the economy—the move from
manufacturing to services and the changing international context.

1.2.1. Inequality and Labour Market Change

Income and wealth are much less equally distributed across the population
than they were 30 years ago. It is rarely understood quite how dramatic that
change has been, nor how important it is for the formulation of public policy
in general and tax policy in particular. The increase in income inequality
over the past 30 years—concentrated in the 1980s—has been without
historical precedent.
In 1978, when the Meade Report was published, someone at the 90th
percentile of the (household) income distribution (in other words, richer
than 90% of the population) had an income three times that of his or her
contemporary at the 10th percentile of the distribution. Now, he or she has
five times as much as that person. In 1978, 7.1 million people had incomes
 Introduction 9

Figure 1.1. Real income growth by percentile point, 1979 to 2009–10 (Great Britain)
Notes: The change in income at the 1st percentile is not shown on this graph. Incomes have been measured
before housing costs have been deducted.
Source: Authors’ calculations using 1979 Family Expenditure Survey and 2009–10 Family Resources Survey.

below 60% of the contemporary median—the person in the middle of the

income distribution from rich to poor. By 2009–10, that figure stood at
13.5 million.6
The change is perhaps best illustrated by a chart such as Figure 1.1, which
shows how real income levels changed at each percentile (i.e. in each
successive 1%) of the overall income distribution between 1979 and 2009–10.
The continuous upward slope of the graph shows that the higher up the
income distribution we go, the higher was the rate of income growth. The
real income of someone at the 5th percentile of the distribution (poorer than
95% of people) was about 30% higher in 2009–10 than that of the equivalent
individual in 1979. Increases were around 50% at the 25th percentile, 70% at
the 75th percentile, and more than 100% at the 95th percentile.
The incomes of the very richest have risen very fast indeed and well away
from those of the rest of the population. Of the 30 million or so people who
pay income tax, about 200,000 are expected to record incomes of over
£200,000 a year in 2011–12, with a further 160,000 having incomes between
£150,000 and £200,000.7 And the richest taxpayers pay a large portion of

6
All figures are measured after housing costs and net of tax payments and benefit receipts, and
are drawn from IFS analysis available at http://www.ifs.org.uk/fiscalfacts.php.
7
HM Revenue and Customs income tax statistics, table 2.5, http://www.hmrc.gov.uk/stats/
income_tax/table2-5.pdf.
10 Tax by Design

total tax revenues: in 2011–12, the top 1% of income tax payers are expected
to pay nearly 28% of all the income tax revenue received by the government,8
more than double the 11% contributed by the richest 1% back in the late
1970s. The poorer half of income tax payers pay just 10% of all income tax.
This extraordinary level of, and increase in, the contribution of the richest is
not down to a more progressive income tax structure—quite the reverse, as
higher rates of income tax are much reduced. Rather, it is down to the very
high levels of income enjoyed by the richest relative to those received by
everyone else.

Table 1.2. Weekly net household incomes in the UK, 2009–10

Mean 10% 25% 50% 75% 90% Percentage

in the UK

Families with children

Lone parent, working £471 £269 £325 £410 £520 £723 4.2%
Lone parent, not working £318 £179 £230 £287 £375 £497 3.9%
Couple, both working £867 £409 £531 £703 £954 £1,356 21.7%
Couple, one working £737 £299 £380 £496 £683 £1,010 11.0%
Couple, not working £390 £130 £256 £347 £465 £610 2.9%

Families without children

Single, working £653 £215 £328 £520 £826 £1,205 11.1%
Single, not working £437 £97 £166 £316 £562 £898 6.4%
Couple, both working £840 £393 £523 £700 £941 £1,312 13.4%
Couple, one working £578 £225 £322 £459 £689 £988 5.4%
Couple, not working £372 £111 £221 £310 £470 £698 2.5%

Pensioners
Single £294 £134 £175 £237 £335 £507 7.2%
Couple £487 £225 £293 £388 £563 £818 10.4%

All £637 £207 £320 £500 £755 £1,098 100.0%

Source: Authors’ calculations based on 2009–10 Family Resources Survey.

8
HM Revenue and Customs income tax statistics, table 2.4, http://www.hmrc.gov.uk/stats/
income_tax/table2-4.pdf.
 Introduction 11

To provide a better understanding of the overall distribution of income,

Table 1.2 provides a detailed snapshot of how net household incomes vary by
family type. It also provides a picture of the distribution of different family
types and of how incomes vary within family types. It illustrates, for
example, the preponderance of two-earner couples, both with and without
children.
Of course, one cannot directly compare the incomes of different family
types to understand relative living standards—couples with children need
more money to live on than the single childless. As one would expect,
couples where both are in work have higher average incomes than any other
group, and single pensioners have lower incomes. But there are big overlaps
between all the groups—no group is universally poor, none is universally
rich. Differences within family types are generally greater than differences
between family types. The tax and welfare systems need to be designed with
all these aspects of the shape of the population and their incomes in mind.
Returning to the question of how the income distribution has become
more dispersed over time, much of this change has resulted from a more
dispersed distribution of wages. In large part, this reflected a rise in the
financial pay-off people received from achieving higher levels of skills and
education, though inequality has also increased dramatically within groups
of people with similar skills. Changes in labour market institutions—for
example, falls in trade union membership—have also played a part. Levels of
unemployment and non-employment also rose rapidly, and whilst official
measures of unemployment fell back from the early 1990s, levels of labour
market participation for men are still well down, reflecting greater numbers
giving illness or disability as the main reason for not working. Over 90% of
working-age men were in employment or self-employment in the mid-
1970s; only 76% were in 2009.9
As we shall see in Chapter 3, and as illustrated in Table 1.3, most of the
reduction in the proportion of men in work occurred among younger and
older age groups. This fact matters enormously in thinking about tax design.
Nobody should argue that the tax and benefit system alone created these
extraordinary changes in participation. Recessions in the early 1980s and

9
Office for National Statistics, Social Trends 2010, figure 4.4, http://www.statistics.gov.uk/
downloads/theme_social/Social-Trends40/ST40_2010_FINAL.pdf.
12 Tax by Design

Table 1.3. Percentages of men and

women in work by age, 1979 and 2008

Age group Men Women

(% in work) (% in work)
1979 2008 1979 2008

16–24 75 58 60 56
25–54 93 88 60 75
55–64 80 67 38 49
All, 16–64 87 78 56 67
Source: Calculations from Labour Force Survey
(with thanks to Antoine Bozio).

early 1990s changed the labour market beyond recognition. But the
incentives in the tax and benefit system can prolong or ameliorate the impact
of such shocks. The recovery after recession in the employment rates of older
workers in the UK is very different, for example, from that in France, where
responses to the recession of the 1980s included generous pension and layoff
arrangements that have not been unwound. Only 19% of men aged 60–64 in
France were in work in 2009 compared with 57% of such men in the UK.10
Another profound change has been in the role of women in the labour
market. Their employment rates have risen—with particularly big increases
among married women with children. As more families with children have,
and aspire to have, both parents working, the impact of tax rates and benefit
withdrawal rates on the potential second earner becomes more important. A
tax system that encourages a primary earner to work by providing tax
credits, which are then withdrawn as family income rises, may discourage a
second earner from working. There is powerful evidence that it is women
with children whose work patterns tend to be most sensitive to the structure
of the tax and benefit system. Making working a bit less attractive relative to
not working tends to have little impact on whether men (at least those aged
25–50) work, but rather more impact on the behaviour of women.11
But there has also been a divergence in the experience of different types of
women in the labour market. While employment rates have increased for

10
Blundell, Bozio, and Laroque, 2011.
11
See Meghir and Phillips (2010) and Brewer, Saez, and Shephard (2010).
 Introduction 13

women in general, they fell to a very low level for the rapidly growing group
of single mothers, recovering somewhat after 1997. The tax and benefit
system has responded to, and also helped to shape, these social changes, with
the introduction and subsequent extension of tax credits doing much to
increase the incentives for single parents to move into work. These sorts of
issues were much less salient on the policy agenda at the time of the Meade
Report in the 1970s. Tax system design needs to be robust to social and
labour market changes. A system that provided generous benefits to non-
working lone parents has, perhaps unsurprisingly, helped facilitate a great
increase in their numbers when economic opportunities for the low-skilled
dried up and social norms changed.
Overall levels of inequality in incomes are also important. In the first place,
of course, we might want to design the tax system to do more work to
ameliorate the underlying growth in inequality. Second, with relatively high
levels of non-employment and low-wage employment, the impact of the tax
and benefit system on both incentives to work and the incomes of low
earners will matter a lot. We come back to this issue in some detail in
Chapter 4. Third, and more generally, a given tax system will have very
different effects depending upon the distribution of incomes among the
population on which it is imposed—the effects of a higher rate of income tax
on earnings over £100,000 are likely to matter more if there are more people
earning over £100,000.
Similarly, the distribution of wealth naturally has consequences for our
views about taxing wealth. If, for example, the distribution of inheritances
were reasonably equal, then the argument put by many for taxing
inheritance on equity grounds would fall away. If, as has indeed happened,
the distribution of wealth and inheritances becomes more unequal, then the
case for a progressive inheritance tax becomes stronger.

1.2.2. Structural Change and Globalization

Changes to incomes and in the labour market themselves, in part, reflect

structural changes in the economy. These have, of course, been considerable
over the period since the publication of the Meade Report. Financial and
14 Tax by Design

business services accounted for 32% of national income in 2008 compared

with just 15% in the late 1970s.12
One very salient example of a change that has mattered enormously to the
tax system has been the increased role of financial services. Their share in
gross value added increased by more than half between 1980 and 2007. This
makes their exemption from VAT an increasingly important tax design
issue—one to which we devote Chapter 8. It also increased the government’s
reliance upon them for corporate and income tax revenues—a fact that has
been painfully underlined in the subsequent financial crisis. The banking,
finance, and insurance sector was responsible for £12.3 billion of corporation
tax in 2007–08—27% of all corporate tax revenues and nearly three times
what might be expected on the basis of its share in the economy. That total
fell to £7.5 billion in 2008–09, just 17% of that year’s corporation tax.13 The
Treasury’s own analysis confirms that the ‘increased importance of the
financial sector’ is one of the factors responsible for ‘increased sensitivity of
receipts to the cycle’.14 It estimates that fully half of the total increase in tax
receipts enjoyed by the government in the five years running up to 2007–08
was due to taxes raised on the housing and financial sectors.15 Unfortunately,
the sensitivity of these sectors to the economic cycle meant that annual tax
receipts from them fell by a full 1½% of GDP16 (over £20 billion) in the two
years from 2007–08 to 2009–10.
The shape of the economy affects the appropriate tax policy. Tax structures
may also help shape the economy, possibly in unwelcome ways. The public
finances suffered as a result of the financial crisis. But the tax system may
itself have played at least some, albeit minor, role in creating or facilitating
the crisis. The tax treatment of housing and financial services, the very low
taxes on capital gains, and the incentives in the corporate tax system for debt

12
Measure refers to gross value added at current basic prices. Data from The Blue Book, ONS.
For the most recent statistics, see table 8.3, The Blue Book, 2010 edition. Time series available
in OECD STAN database (variable id ‘VALU’).
13
Source: HM Revenue and Customs corporation tax statistics, tables 11.4 and 11.5,
http://www.hmrc.gov.uk/stats/corporate_tax/menu.htm.
14
HM Treasury, 2008, 17.
15
Including corporate taxes, stamp duties, and income tax and National Insurance raised from
earnings from financial sector employees.
16
Source: HM Treasury, 2010a, box C3.
 Introduction 15

funding over equity funding are all possible culprits. We address all these
issues in the relevant chapters.
Meanwhile, technological advances have transformed the productive
economy from which taxes are raised, while at the same time making it
easier to administer the tax system and easier to structure activities to avoid
tax. The role of technology in cross-border transactions and the implications
of this for the tax system are touched on in Chapters 7 and 18. This and
other aspects of globalization mean that cross-border issues more generally
are now much more important than they were 30 years ago, and this has
undoubtedly placed new constraints on what is possible within the tax
system. Real complexities are also created by increased globalization—for
example, regarding the appropriate treatment of companies by national tax
systems. How does one think of taxing a Swiss company that develops a drug
in a research facility in the UK, manufactures it in Belgium, and mainly sells
it in the US?17
But despite some predictions to the contrary, countries are not being
forced inexorably to tax less in an increasingly globalized and competitive
world economy. Between 1975 and 2008, taxes rose as a proportion of
national income in virtually every OECD country. On average, the tax take
rose from 29.4% to 34.8% of national income. In no OECD country was
there a significant fall in the tax take over this period. And the variation
between countries is striking. Denmark, Sweden, the US, and Japan are all
rich countries. In Denmark and Sweden, taxes accounted for 48% and 46%
of GDP respectively in 2008. In the US and Japan, they accounted for only
26% and 28%.18 There is no straightforward relationship between the total
tax burden and economic performance.
It clearly remains possible for a successful economy to raise 40% or more
of national income in tax despite the pressures of globalization. But, within
the total tax take, we might expect that governments would find it more
difficult to raise taxes from internationally mobile companies and people. In
fact, revenue from corporation taxes has more than held up over the past 40
years—corporate income taxes accounted for 9% of tax revenues across the

17
Example used by John Kay at a conference at the London School of Economics in May 2011.
18
All figures from OECD tax revenue statistics, table A, http://www.oecd.org/document/60/
0,3746,en_2649_34533_1942460_1_1_1_1,00.html#A_RevenueStatistics.
16 Tax by Design

OECD in 1965, 8% in 1985, and 10% in 2008.19 As we have already seen, the
richest—and probably most mobile—1% of taxpayers in the UK are expected
to contribute 28% of income tax revenue in 2011–12, compared with just
11% in 1979.
This is not to imply that there are no problems. The fact that some people
and companies may be able to avoid taxes because of their international
mobility might mean that the burden is in some respects now borne more
unequally, and inefficiently, than before. In addition, if there is a sense that
some individuals or companies can avoid paying tax because of their
domicile or ability to shift profits around, then acceptance of the system and
belief in its equity may be damaged.
However, the resilience of these revenues highlights the fact that while it
has certainly become easier and cheaper to cross national borders—and it
may well continue to become so—it is far from costless. Policymakers must
therefore decide whether to design the tax system in readiness for the day
when globalization does indeed make it much harder to raise revenue from
mobile individuals and businesses, or whether to collect the revenue while
they can and reform the system once these pressures have materialized.
Either way, for the time being, as we discuss in Chapters 4 and 18,
globalization certainly affects the rates at which we can reasonably hope to
tax high incomes and the feasible structure of corporate taxes.

1.3. THE POLITICS OF TAX REFORM

Of course, it is not just the economic context that matters for tax policy.
Politics matters too. It is not possible to understand the structure or
development of the tax system outside of a political context—not least
because making people obviously, or apparently, worse off is rarely good
politics, at least in the short term. And almost all tax reforms make some
people worse off.
This makes much reform rather harder to put into practice than to design.
Worse, when governments need money they tend to look to raise revenue in

19
Source: OECD tax revenue statistics, table C, http://www.oecd.org/document/60/0,3746,en_
2649_34533_1942460_1_1_1_1,00.html#A_RevenueStatistics.
 Introduction 17

ways that make the losers relatively hard to identify. This results in
complexity and poor policy. A classic example is the way in which
governments have often used fiscal drag to increase income tax revenues
whilst cutting the basic rate of income tax. Fiscal drag occurs where tax
bands and allowances are raised more slowly than the incomes on which the
tax is levied. Fiscal drag was largely responsible for an increase in the
number of income tax payers from fewer than 26 million in 1996–97 to a
peak of 32.5 million in 2007–08.20 The number paying the higher (40%) rate
of income tax roughly doubled from 2 million to nearly 4 million over the
same period. These are big changes that mean that some people have gained
less from higher wages and salaries than they would otherwise have done—a
fact that appears to get little play in the policy debate.
For similar reasons, rates of National Insurance contributions have risen
while income tax rates have fallen. Compared with income tax, NI taxes a
narrower range of income, does less to redistribute resources from rich to
poor, and is less transparent to the citizen, especially that part which is levied
formally on employers. As we will see in Chapter 4, its original function as a
payment for rights to contributory benefits such as the state pension has
been almost entirely eroded, although governments exploit the lingering
belief that the link is still a strong one. Indeed, a remarkable number of
people believe that NI in some way pays for the National Health Service—a
misapprehension played upon in 2002 when an increase in NI rates was
announced purportedly to pay for higher spending on the NHS.
Raising revenue through more radical reform has proved difficult. The
attempt in the early 1990s to impose VAT at the full rate on domestic energy
consumption created a political backlash strong enough to see the policy
partially abandoned, and then further reversed by the next government. This
reform was to have been accompanied by measures that left most poorer
people better off, while still raising revenue overall. But people who spend
unusually large proportions of their income on energy would still have been
left worse off. That makes the change difficult to achieve in political terms.
But suppose we started, as some countries do, in a world where VAT was
already levied on fuel. To abolish it would not look like an attractive policy.
The rich spend more in absolute terms on fuel than do the poor. Hence,

20
HM Revenue and Customs income tax statistics, table 2.1, http://www.hmrc.gov.uk/stats/
income_tax/table2-1.pdf.
18 Tax by Design

abolishing an existing tax would look like subsidizing the well-off to increase
their consumption of a polluting good. But we rarely think of the fact that
failing to tax something is, in effect, subsidizing its usage. This thought
experiment is also valuable in demonstrating the extent to which tax is one
area of public policy where the ‘tyranny of the status quo’ is strongest.
Changing it substantively is difficult. We hope to challenge some of that
tyranny here.
Whilst the attempt to place full VAT on domestic energy consumption was
unsuccessful, weakness in the public finances can sometimes facilitate
beneficial tax reform. If people accept that there is a need to raise money,
then there is no escaping the need to leave some people worse off. The period
after 1992—the last significant episode of fiscal consolidation—saw the final
phasing-out of mortgage interest relief and the married couple’s allowance as
well as big increases in petrol and tobacco duty, policies that might have
been hard to implement in the absence of a widely recognized need for fiscal
policy to be tightened. The recent crisis has seen less reform—increases to
rates of NI, to higher rates of income tax, and to the VAT rate have been
used to raise money. That is something of a shame. As Rahm Emanuel
observed when he was President Obama’s Chief of Staff, one should never let
a crisis go to waste.
Some poor policymaking can be understood, if not excused, by reference
to straightforward political pressures. The complexity of the tax system and
lack of public understanding make poor policy, and indeed explicit
misrepresentation by government, much too easy. They can lead to an
undesirable narrowing of public debate and a fixation on a few easy-to-
understand elements, such as the basic rate of income tax. This then drives
out wider considerations. Governments have also been accused of relying on
so-called ‘stealth taxes’, taxes where the incidence on individuals is unclear.
The use of such taxes can itself undermine trust in the tax system.
The significant and growing complexity of business taxation makes public
debate in this area very limited indeed. Corporation tax and business rates
between them raise almost as much as VAT, nearly half as much as income
tax, and nearly three times as much as fuel duties. And, of course, they must
all be paid by individuals in the end. Their lack of salience and the lack of
debate are regrettable. The increasing complexity of corporate taxes also
leaves room for a great deal of corporate lobbying to introduce and extend
 Introduction 19

special treatments and allowances. The story of the R&D tax credit is a good
example, as commented on by Alt, Preston, and Sibieta (2010, 1205):
Enacting tax policy can create interest groups and constituencies in favour of that
policy. Even when they did not lobby for the policy in the first place, … they will
lobby both for persistence and extensions that allow policy to drift from its original
motivation. Therefore, any potential tax reformer should remember that any new
allowances enacted or favourable tax treatments provided to particular groups could
prove difficult to remove and may be distorted into something different over time.
Good tax policy requires an open, transparent, and well-informed public
debate based on credible data. Poor public understanding is a constraint on
good tax policy. It allows poor-quality analyses of policy reforms to gain
prominence.
Good tax policy also requires effective processes within government. At
present in the UK, there is arguably a more limited level of discussion and
debate about tax policy within government, and as part of the legislative
process, than in other areas of policy. The Treasury is a remarkably powerful
institution and, as far as tax policymaking is concerned, has become more
powerful in recent years, as it has taken on much of the policymaking
capability of HMRC. There are no checks and balances within the executive.
The Chancellor effectively takes sole responsibility for his Budget.21 The
legislature—parliament—also effectively has a rather weak oversight role,
particularly when it comes to some of the more complex areas of tax policy.
Finally, we should not forget one very important change to the political
and institutional environment for tax reform over the past 30 years: the
growing integration of the European economies and the increased influence
of the European Union (EU). Thirty years ago, the main role of the
European Union (then the European Economic Community, EEC) was to
bring down trade barriers. Since then, it has grown in importance and it now
exerts an important influence on UK tax policy, both directly and indirectly.
Member states of the EU retain sovereignty over direct taxation, but are
nevertheless required to exercise it in accordance with EU law. In recent
years, there have been a number of successful legal challenges to elements of
national corporate income taxes at the European Court of Justice (ECJ). This
has prompted a variety of reforms—for example, changes to anti-avoidance

21
In the UK, it has always been his Budget. We have had women Foreign Secretaries and
Home Secretaries and most famously a woman Prime Minister, but never a female Chancellor.
20 Tax by Design

rules designed to limit the ability of multinational firms to shift their taxable
income between countries with different tax rates.
The EU has greater formal influence over indirect tax policy, including
explicit limits on the ability of countries to alter their VAT rates. This reflects
the fact that an open market in goods requires some kind of system for
dealing with differential tax treatment of goods across country borders.
Issues of dealing with VAT and excises within the EU are central to our
discussion in Chapter 7.

1.4. CONCLUSIONS

Taxes, like death, are unavoidable. But we can design our taxes. We are not
bound to have a tax system as inefficient, complex, and unfair as our current
one. To improve things, we need to see the system as a whole, we need to
design the system with a clear understanding of the population and economy
on which it operates, and we need to apply economic insights and evidence
to the design. We also need a much more informed public debate and a
much better set of political processes than the ones we currently have.
Our purpose in this book is to create a framework and directions for
reform. In the next chapter, we explain in more depth what we mean by an
economic approach. We then look in turn at taxes on earnings, taxes on
consumption, environmental taxes, taxes on savings, taxes on wealth, taxes
on land and property, and taxes on companies. In the final chapter, we draw
all this together to propose an overall set of reforms to the tax system which,
we believe, has the potential to make us all better off and free us from at least
some of the shackles created by the complexities and inefficiencies of the
current system.