THE IFS GREEN BUDGET:
JANUARY 2003
Robert Chote
Carl Emmerson
Helen Simpson
supported by
E •S •R •C
ECONOMIC
& SOCIAL
RESEARCH
COUNCIL
THE INSTITUTE FOR FISCAL STUDIES
Commentary 92
Contents
List of figures
List of tables
v
vii
1
Summary
2
2.1
2.2
2.3
2.4
Planning the public finances
The Chancellor’s fiscal rules
Planning and forecasting revenues
Planning government spending
Conclusions
7
7
20
25
29
3
3.1
3.2
3.3
3.4
IFS public finance forecasts
Borrowing in 2002–03
Borrowing in 2003–04
Medium-term prospects
The Budget judgement
31
32
34
35
39
4
42
4.4
4.5
What do the child poverty targets mean for the child tax
credit?
Child poverty under Labour
The child tax credit
Is the government likely to meet its child poverty target in
2004–05?
What would it cost to reduce child poverty further?
Conclusion
5
5.1
5.2
5.3
Income tax and National Insurance contributions
Income tax and National Insurance: a history of convergence
The system from April 2003
Options for further reform
54
54
57
60
6
6.1
6.2
6.3
6.4
6.5
Company taxation and innovation policy
The August 2002 consultation
Dividend taxes
North Sea taxation
Stamp duty on shares
R&D and policy towards innovation
64
64
68
71
74
75
7
7.1
7.2
7.3
7.4
Childcare subsidies
Why subsidise childcare?
Families’ employment and childcare use
Options for subsidising childcare
Potential numbers of eligible families and budget costs,
holding childcare and employment constant
Allowing childcare responses (holding employment constant)
Considering employment responses
Impacts across families
Conclusions
85
86
87
89
92
4.1
4.2
4.3
7.5
7.6
7.7
7.8
1
42
44
48
51
53
95
100
101
102
8
8.1
8.2
Measuring public sector efficiency
Efficiency measurement in the public sector
Conclusions
105
106
110
9
9.1
9.2
The distributional effects of fiscal reforms since 1997
What fiscal reforms?
The distributional impact of reforms directly affecting
households
Limitations of the distributional analysis
Taxes on business
Stamp duty on residential properties
Conclusion
112
112
115
Appendix A: Forecasting public finances
Appendix B: Distributional effects of pre-announced direct
tax and benefit reforms due in 2003–04
Appendix C: Budgets since 1979
Appendix D: Headline tax and benefit rates and thresholds
Appendix E: Tax revenues ready reckoner
129
135
9.3
9.4
9.5
9.6
117
119
124
128
142
147
150
Figures
Figure 2.1
Figure 2.2
Figure 2.3
Figure 2.4
Figure 2.5
HM Treasury estimates of the output gap
Current budget surplus as a percentage of national income
Public sector net debt as a percentage of national income
Public sector net debt as a percentage of national income
Estimated future payments under Private Finance Initiative contracts as a
percentage of national income
Revenues from stamp duty on shares and property
Nominal growth in national income, the FTSE All-Share Index and house prices
Non-North-Sea corporation tax revenues and forecasts as a share of national
income
Public sector net investment: out-turns and forecasts as a percentage of national
income
8
9
12
15
16
Non-North-Sea corporation tax receipts and forecasts as a percentage of national
income
Current budget surplus as a percentage of national income
36
Figure 4.1
Figure 4.2
Figure 4.3
Figure 4.4
Child poverty (children in households with less than 60% median income AHC)
Financial support for a family with one child under the outgoing system
Financial support for a family with one child under the new tax credits
Distributional effects of different increases in child-related benefits and tax
credits in April 2004
44
46
47
52
Figure 5.1
Figure 5.2
Combined payroll tax schedule, 2003–04
Losses across the income distribution from National Insurance changes
announced in the 2002 Budget, and from a package of similar changes to income
tax
58
60
Figure 6.1
Figure 6.2
Figure 6.3
Figure 6.4
GERD as a percentage of GDP: G5 countries
BERD as a percentage of GDP: G5 countries
UK R&D levels: BERD and R&D Scoreboard
UK BERD as a percentage of GDP: industry breakdown
76
78
78
79
Figure 8.1
SFA and DEA
108
Figure 9.1
Figure 9.2
Figure 9.3
Impact of direct personal tax, benefit and expenditure tax changes since 1997
Impact of direct personal tax, benefit and expenditure tax changes since 1997
Impact of direct personal tax, benefit and expenditure tax changes since 1997,
with and without effect of stamp duty
116
118
127
Figure 2.6
Figure 2.7
Figure 2.8
Figure 2.9
Figure 3.1
Figure 3.2
22
22
25
27
40
Tables
Table 2.1
Average errors in forecasting public sector net borrowing, as a percentage of
national income and in £ billion
Forecast capital spending by the public sector under both conventional finance
and the Private Finance Initiative
Public finances across the EU in 2001, percentage of national income
The size of the AME margin
13
Comparison of Green Budget and HM Treasury forecasts for government
borrowing, 2002–03
Comparison of Green Budget and HM Treasury forecasts for government
borrowing, 2002–03 and 2003–04
Medium-term public finances forecasts, based on cautious macroeconomic
assumptions
Medium-term public finances forecasts, based on cautious macroeconomic
assumptions
32
Table 4.1
Table 4.2
Possible changes in child poverty (AHC), 2000–01 to 2004–05
Effect of possible increases in per-child element of the child tax credit in April
2004
50
51
Table 5.1
Rates of income tax and National Insurance contributions, 2003–04
57
Table 6.1
Table 6.2
Table 6.3
Applicable taxes and marginal tax rates by date of approval of field
Breakdown of R&D by who performs it and who funds it
Percentages of BERD by who funds it
72
77
77
Table 7.1
Table 7.2
Table 7.3
Table 7.4
Employment patterns among families with children
Childcare use for families with no parent working less than 16 hours each week
Number of eligible families and costs of the childcare credit: childcare unchanged
Number of eligible families and cost of the childcare credit: with potential
childcare responses
87
88
93
98
Table 9.1
Table 9.2
Table 9.3
Table 9.4
Table 9.5
Revenue effects in 2003–04 of changes to taxes and benefits made since 1997
Revenue effects of major changes in ‘business taxes’
Rate of stamp duty on property, 1997 to present day
Yield of stamp duty on residential property, 2001–02
Distribution of the value of homes worth more than £250,000, across the income
distribution
113
122
124
125
126
Table A.1
130
Table A.2
Table A.3
Table A.4
A comparison of last year’s Green Budget forecast and the Treasury November
2001 Pre-Budget Report forecast with the estimated out-turn for 2001–02 from
the November 2002 Pre-Budget Report
IFS Green Budget and Treasury main errors in forecasting tax receipts, 2001–02
Forecasts for government borrowing in 2002–03
Main macroeconomic assumptions used in the baseline forecast
Table B.1
Table B.2
Table B.3
Table B.4
Table B.5
Table B.6
Table B.7
Percentage gains across the income distribution from reforms in 2003–04
Weekly cash gains across the income distribution from reforms in 2003–04
Percentage gains from reforms in 2003–04, by family type
Weekly cash gains from reforms in 2003–04, by family type
Numbers of winners and losers, by family type
Income bands for each decile for different family types
Estimated number of families in Great Britain, by family type
138
138
139
139
140
141
141
Table E.1
Direct effects of illustrative changes in taxation to take effect April 2003
150
Table 2.2
Table 2.3
Table 2.4
Table 3.1
Table 3.2
Table 3.3
Table 3.4
14
19
29
33
37
38
130
132
134
1. Summary
Planning the public finances
The Chancellor’s tax and spending decisions are constrained by two selfimposed rules. The golden rule states that borrowing should only pay for
investment, while the sustainable investment rule limits public debt to no more
than 40% of national income. Both have to be satisfied over the economic
cycle, but not every year.
The Treasury judges progress against the golden rule by looking at the average
surplus on the current budget (which excludes investment spending) since the
beginning of the present economic cycle in 1999–2000. Large surpluses in the
early years mean the golden rule is likely to be overachieved comfortably in
the current cycle. But more important in framing the Budget should be
whether the present stance of fiscal policy is consistent with meeting the
golden rule looking forward.
In the past, the Chancellor has sought to overachieve the golden rule by
around 0.7% of national income. Treasury estimates suggest that this is
enough to ensure that the golden rule would still be met if the trend level of
economic activity consistent with stable inflation had been overestimated by
1%. In the 2002 Pre-Budget Report, the Chancellor forecast that the cyclically
adjusted surplus would return to this level by 2007-08, but this prediction may
be unduly reliant on factors such as ambitious forecasts for corporation tax
receipts.
The sustainable debt rule is not currently as binding as the golden rule, with
the Treasury expecting the debt-to-GDP ratio to rise only fractionally to 33%
by 2007–08. The debt measure used does not include all government liabilities
– for example, some arising from the Private Finance Initiative. We estimate
that if all capital spending under PFI deals signed to date were funded
conventionally, public sector net debt would be 3.8% of national income
higher by March 2006.
If the UK were to join the Euro, fiscal policy might be further constrained by
the Stability and Growth Pact and the Excessive Deficits Procedure in the
Maastricht Treaty. The former requires member countries to aim in the
medium term for a budget balance ‘close to balance or surplus’. The latter
requires general government gross debt below 60% of GDP and a budget
deficit of less than 3% of GDP (unless it is both exceptional and temporary or
it is demonstrably declining towards these levels). The golden rule is by no
means perfect, but a balanced-budget rule would make it more difficult to
carry out investment that benefits future generations.
But the Stability and Growth Pact might be interpreted in a less restrictive way
by the time the UK joins the Euro – if it joins at all. The European
Commission has suggested that the balanced-budget rule should be assessed
with reference to where the economy stands in the economic cycle and that it
should have scope to spread the cost of beneficial investment and tax reforms
that raise employment or growth potential across generations. This would
1
Green Budget, January 2003
bring the Stability and Growth Pact slightly closer to the UK rules. In any
event, the UK’s fiscal position is relatively strong in comparison with those of
most Eurozone countries.
IFS public finance forecasts
In the Pre-Budget Report (PBR) last November, the Chancellor conceded that
the unexpected weakness of the economy was depressing tax revenues and
that this would force him to borrow more than he had expected this year and
next. But over the medium term, he predicted that the economy would bounce
back to its trend and that the public finances would return pretty much to the
path he had expected in last April’s Budget.
In the short run, our forecasts are very similar to those in the PBR. In 2002–
03, we forecast public sector net borrowing of £22.1 billion, slightly higher
than the PBR forecast of £20.1 billion. In 2003–04, we forecast public sector
net borrowing of £25.2 billion, again slightly higher than the £24.5 billion
forecast in the PBR. We expect deficits on the current budget of £8.8 billion
this year (compared with the Treasury’s £5.7 billion) and £5.1 billion next
year (compared with the Treasury’s £4.9 billion).
But in the medium term, we believe the public finances will be weaker than
the PBR suggested in November, even if the economy behaves much as the
Treasury expects. In part, this is because we do not expect that the loss of
revenue that the Treasury attributes to the stock market and the plight of
financial companies will recover as sharply as the Chancellor predicts. Hence,
for example, we are less optimistic about the medium-term path of corporation
tax revenues. By 2005–06, we forecast public sector net borrowing of
£28 billion, compared with the PBR forecast of £19 billion.
Nonetheless, we believe the Chancellor can credibly claim that both the
golden rule and the sustainable investment rule will have been met
comfortably over the current economic cycle, which is projected to end in
2005–06. We expect the government to overachieve the golden rule over the
current cycle by around £31 billion, compared with the £46 billion the
Chancellor predicted in November.
But looking forward into the next cycle, our forecasts imply that spending cuts
or tax increases will be required if the golden rule is to continue to be expected
to be met. If the Chancellor sticks by his PBR forecasts, then he might well
argue that no fiscal tightening is necessary. But if our forecasts are correct, he
would be unlikely to be able to avoid such measures for long without
undermining the credibility of the fiscal rules that he has set such store by. In
that event, spending cuts would sit oddly with both the government’s stated
objectives and its previous actions. Tax increases would, therefore, seem more
likely.
The tax increases required depend on how cautious the government wants to
be. To expect to meet the golden rule exactly would, we estimate, require tax
increases of around £4 billion. But in the past, the Chancellor has been more
cautious and sought to overachieve the golden rule by around 0.7% of national
income on average. To expect to do this would require tax increases of around
£11 billion.
2
Summary
What do the child poverty targets mean for child tax credit?
The government has pledged to eradicate child poverty within a generation. As
an intermediate target, it has also set itself the goal of reducing the number of
children in poverty by at least a quarter by 2004. In practice, this means
reducing by a quarter the number of children in households that have incomes
below 60% of the median, which can be measured either before or after
housing costs. Like the government, we focus on the latter measure. There
were 4.2 million children in poverty on this definition in the base year of
1998–99, which implies a target of around 3.1 million in 2004–05.
By 2000–01 (the last year for which we have data), the number of children in
poverty on this definition had fallen to 3.9 million. So if the government is to
hit its target, child poverty must fall by a further 0.8 million. This is an
average of 200,000 a year, which is faster than it has fallen to date. Whether
this will happen depends not only on the financial support that the government
offers low-income families, but also on the impact of economic and
demographic factors on their incomes and on the median income against
which they are compared.
The child tax credit, to be introduced in April 2003, will represent the majority
of government support for children and is the government’s main instrument
for targeting child poverty. Our best estimate is that existing tax and benefit
reforms will take 800,000 children out of poverty between 2000–01 and 2004–
05, but that the rise in the median income from earnings growth will ‘move the
goalposts’ and put 200,000 of them back in. This estimate implies that the
government would fall 200,000 short of its target on the after-housing-costs
measure, although there are considerable uncertainties around this prediction.
(We assume that population, employment rates and household composition do
not change, and that average earnings rise in line with recent trends.)
If our estimates are correct, the government could argue that it is on course to
hit its target on the before-housing-costs measure of poverty. But what could it
do to expect to meet the target on the after-housing-costs measure? We
estimate that taking a further 200,000 children out of poverty could be
achieved by raising the per-child element of the child tax credit by £3 per
week (on top of the increase in line with average earnings that has already
been promised) at a cost of £1 billion. Out of the options that we consider, this
is the most cost-effective, as it is the best-targeted at reducing poverty.
Income tax and National Insurance contributions
The income tax and National Insurance systems in the UK have become
steadily more similar over the past 40 years. Gordon Brown’s decision to raise
National Insurance from April – in part by requiring an employee contribution
for the first time on all earnings above the upper earnings limit of £30,940 a
year – is another step in this direction.
Labour promised in the 1997 election campaign, and again in 2001, not to
raise the basic or top rates of income tax. But the distributional impact of
April’s increases in employer and employee National Insurance contribution
rates will be similar to a 2p increase in the lower, basic and higher rates of
income tax. Differences arise because income tax is levied on income from
3
Green Budget, January 2003
sources other than earnings, whereas National Insurance is paid only on
earnings. The self-employed are also unaffected by the rise in the employer
contribution rate.
Given the erosion of the ‘contributory principle’ originally used to justify
National Insurance – namely, that people’s social benefit entitlements were
meaningfully linked to what they paid in – there now seems little economic
justification for separate income tax and National Insurance systems. But this,
and previous, governments appear to believe – for now at least – that voters
are happier to pay more National Insurance than more income tax.
This suggests that further gradual alignment of the two systems might be more
likely than wholesale integration. Closing the gap between the earnings level
at which employee National Insurance contributions drop from 11% to 1%
(£30,940) and the income level at which income tax rises from 22% to 40%
(£35,115) is an obvious next step and would also raise significant extra
revenues for the Treasury. Revenue could also be raised by further increases in
the rates of employer or employee National Insurance. The latter could be
done solely on earnings above £30,940 if the Chancellor wanted to raise
revenue in a very progressive way.
Company taxation and innovation policy
Last August, the government issued a consultation document on further reform
to the corporation tax system. The objective of the proposals is to align the
calculation of taxable profits more closely with the measurement of profits in
company accounts. But replacing the current system of capital allowances
with a deduction for the depreciation charge in accounts could see big winners
and losers among different sectors. Other proposals would have implications
for the extent to which companies could offset losses in one part of their
business against profits elsewhere.
In 1997, the UK government increased the taxation of dividend income for
pension funds and some other institutional investors, with the goal of boosting
investment. The USA is now proposing to reduce the taxation of dividends
with partly the same objective. It is doubtful that the changes in either country
will have much impact on investment or share prices. But the US reform
would benefit US citizens who hold shares directly, with the wealthiest
gaining most.
The changes to North Sea taxation announced in the 2002 Budget and PreBudget Report raise extra revenue from this sector. They provide some
welcome simplification and are in line with the principles of efficient resource
taxation. But they could have gone further. The rules have been changed
frequently in the past in response to revenue needs and changes in the oil
price. A stable tax regime that could cope automatically with changing oil
prices would help investment planning in the sector.
Cutting stamp duty on share transactions would reduce distortions and may
contribute to the Chancellor’s goal of boosting UK productivity. Stamp duty
lacks investment allowances, and is therefore more likely to deter investment
than other capital taxes. It reduces the efficiency of the stock market by raising
transaction costs and may increase share price volatility. It also distorts merger
4
Summary
and acquisition activity, producing a bias towards overseas rather than UK
ownership of companies.
Childcare subsidies
Over the last decade, several initiatives have been undertaken to help families
with the costs of childcare, both on distributional grounds and to encourage
parents into paid work. Among them is the childcare credit in the working
families’ tax credit, which will be transferred to the new working tax credit in
April. The Chancellor has said that he wants to do more to help ‘parents to
make real and effective choices on balancing work and family life’. Chapter 7
considers ways of extending the scope and generosity of the childcare credit.
Options for reform might include: increasing the proportion of childcare costs
that can be claimed; raising the cash ceiling on claims; extending the subsidy
to ‘informal’ childcare provided by family and friends; allowing parents to
claim if they work less than 16 hours a week; and making the means-testing of
the credits more generous.
Costing these reforms is difficult, as it is hard to predict how many parents
would take up an expanded credit and to what extent they would change their
employment behaviour and use of paid childcare. Covering a greater
proportion of formal childcare costs or extending coverage to informal care
could be very expensive if the change encouraged much greater use of
subsidised childcare. There is also a danger that parents could claim more
money from the government without any significant increase in childcare use.
This suggests that any expansion of the childcare credit would have to be
designed in such a way as to promote value for money. Such options could
include: lowering the weekly ceilings on subsidised childcare spending;
specifying the value of the credit per hour of childcare; linking the maximum
subsidy to the main carer’s working hours; or linking eligibility to the
employment and earnings of the main carer.
Measuring public sector efficiency
The government has staked its political credibility on delivering clear
improvements in public services. Substantially increased resources are being
ploughed into areas such as health, education and transport. The challenge for
spending departments now is to ensure that these resources are used as
efficiently as possible. To help achieve these improvements, the 2002
Spending Review set out performance targets in around 130 Public Service
Agreements, covering both outcomes in particular services and the efficiency
with which these outcomes are achieved.
In April 2000, the Treasury’s Public Services Productivity Panel proposed a
new approach to measuring the efficiency of police authorities, drawing on
two techniques: stochastic frontier analysis and data envelopment analysis.
The government has also commissioned research into their possible use to
assess local authorities.
The results of applying these efficiency measurement techniques are sensitive
to a number of factors, including: which inputs and outputs are considered;
5
Green Budget, January 2003
errors in measuring these inputs and outputs; and the statistical assumptions
made. While these methods are potentially useful, the results should be treated
cautiously, as should their application as part of a system to affect providers’
performance incentives.
The distributional effects of fiscal reforms since 1997
The government has carried out many reforms to the tax and benefit system
since coming to office in May 1997. Comparing the system as it will operate
next year with that which Labour inherited (adjusted for inflation and statutory
uprating) suggests that there have been around £51.7 billion of revenue-raising
measures and £53.3 billion of revenue-reducing measures. This leaves a net
‘giveaway’ of around £1.6 billion, or about £1.50 per household per week.
Analysing the distributional impact of such changes, IFS research traditionally
focuses on measures that directly affect household incomes and spending. This
shows a progressive pattern, varying from a boost of more than 15% to the
incomes of the poorest tenth of the population to a loss of nearly 3% for the
richest tenth. The average impact is an increase in income of around 2%.
But focusing on those tax and benefit changes that are relatively easy to
allocate to individual households shows a much more generous net ‘giveaway’
than taking all tax and benefit changes into account. Ultimately, all taxes have
to be paid by individuals and including those that we do not model would
reduce incomes on average by 1.7%, wiping out much of the average net gain.
Taxes formally on business are one category that is excluded from our
traditional distributional analysis. But assessing what ‘taxes on business’ are –
and whether they have risen – is not easy. If we focus on taxes levied on
company profits, the net impact of Labour’s reforms since 1997 is an
estimated tax increase of around £4 billion in 2002–03, falling to an estimated
£1.4 billion in the next year, as the transition to quarterly payments of
corporation tax ends. Extending our distributional analysis by attributing this
tax increase to households owning shares – who tend to be on relatively high
incomes – might make Labour’s reforms look even more progressive. We
estimate that the same is true of the increases Labour has implemented for
stamp duty on house purchases since 1997.
6
2. Planning the public finances
This chapter of the Green Budget discusses the planning of the public finances
in the context of the fiscal rules that the government set for itself in 1998.
Section 2.1 explains the rules and how compliance with them is assessed and
looks at the errors made in the past when forecasting the public finances. It
then examines whether getting the private sector to finance public investment
can circumvent the rules. It concludes by comparing the current rules with the
Growth and Stability Pact that would apply if Britain joined the Euro. Section
2.2 examines current issues that arise in planning and forecasting government
revenues. It focuses on the effect of movements in the economic cycle and
asset prices – with particular reference to stamp duty and corporation tax.
Section 2.3 turns to current issues in forecasting and planning government
spending. It focuses on public investment, public sector pay and the safety
margin the government provides itself for unexpected spending demands.
2.1 The Chancellor’s fiscal rules
In July 1998, the Chancellor outlined two fiscal rules that would constrain his
tax and spending decisions. The stated rationale was to provide a credible
framework within which the government could operate, to ensure the
sustainability and fairness of the public finances.
•
The golden rule states that the government will only borrow to fund
investment and not current expenditures. This aims to ensure that future
generations will only be repaying debt the accumulation of which benefits
them through the stock of capital it financed.
•
The sustainable investment rule states that public sector debt should
remain at a ‘stable and prudent’ level, interpreted by the Chancellor as no
more than 40% of national income. By constraining the total level of debt
allowable, the long-term sustainability of the public finances is ensured.1
These rules have to be met over the ups and downs of the economic cycle,
rather than each and every year. When activity in the economy runs below the
trend level thought consistent with stable inflation, weaker profits and higher
unemployment mean that tax receipts are lower and government spending
higher than can be sustained over the long term. The resulting injection of
spending power into the economy in itself helps take activity back up to a
sustainable level. Conversely, when activity is above trend, temporarily
buoyant tax receipts and lower social security costs take spending power out
of the economy and help cool it down. Applying the fiscal rules over the cycle
rather than in every year allows these ‘automatic stabilisers’ to operate
1
For a more in-depth discussion of the sustainability of the public finances, see HM Treasury,
Long-Term Public Finance Report: An Analysis of Fiscal Sustainability, London, 2002
(www.hmtreasury.gov.uk/Pre_Budget_Report/prebud_pbr02/assoc_docs/prebud_pbr02_adsustain.cfm).
7
Green Budget, January 2003
unimpeded.2 The sustainability of the public finances is ensured since extra
borrowing undertaken because the economy is below trend has to be repaid
when it moves above trend.3
Assessing whether the golden rule is met
The golden rule stipulates that the government’s borrowing should not exceed
its spending on investment over the economic cycle. This means that tax
revenue has to be sufficient, on average, to pay for the government’s current
spending (including depreciation). In other words, the ‘current budget’ – the
surplus of receipts over current spending – has to be in balance or in surplus
on average over the cycle.
Whether the golden rule has been met can only be judged definitively in
retrospect, by examining the behaviour of the current budget over a full
economic cycle. But it is hard to judge precisely where the economy is in the
cycle at any given time. To do so, it is necessary to estimate the ‘output gap’ –
a measure of where national income stands relative to the trend level assumed
consistent with stable inflation.
Figure 2.1. HM Treasury estimates of the output gap
4
Percentage of trend output
3
2
1
0
-1
-2
-3
-4
1990
1992
1994
1996
1998
2000
2002
2004
2006
Year
Note: Actual output less trend output as a percentage of trend output (non-oil basis).
Source: Chart A4 of HM Treasury, Pre-Budget Report: 2002, Cm. 5664, London, 2002
(www.hm-treasury.gov.uk/Pre_Budget_Report/prebud_pbr02/prebud_pbr02_index.cfm).
The Treasury’s estimates of the output gap from 1990 to 2008 are shown in
Figure 2.1. The current cycle is assumed to have begun when output moved
2
The automatic stabilisers operate in the right direction, but the strength with which they do
so is a function of the precise structure of the tax and benefits system and may not be optimal
from the perspective of macroeconomic management.
3
A more detailed discussion of the government’s fiscal rules can be found in C. Emmerson
and C. Frayne, The Government’s Fiscal Rules, Briefing Note no. 16, IFS, London, 2001
(www.ifs.org.uk/public/bn16.pdf).
8
Planning the public finances
above trend in 1999–2000 following a mini-cycle between quarter 1 of 1997
and mid-1999. Output then fell below trend again in 2001–02. Stronger
growth is forecast to close the negative output gap by 2005–06, bringing the
current cycle to an end.4
Figure 2.2 shows the surplus on current budget from 1966–67 to 2007–08,
according to HM Treasury forecasts. Since the introduction of the golden rule,
the current budget has been in surplus every year. This is in contrast to most of
the years from 1973–74 until 1997–98, when current budget deficits occurred
in all but the three years from April 1988 to March 1991. Over the cycle
running from 1986 to 1997, the golden rule was far from met. Current budget
deficits averaging over 4% of national income between 1991–92 and 1996–97
far outweighed the modest surpluses in the previous years.
Figure 2.2. Current budget surplus as a percentage of national income
8
Percentage of GDP
6
4
Actual
HM Treasury November 2002 Pre-Budget Report forecast
Cyclically adjusted surplus
2
0
-2
-4
-6
-8
66–67 69–70 72–73 75–76 78–79 81–82 84–85 87–88 90–91 93–94 96–97 99–00 02–03 05–06
Financial year
Note: Measures exclude the windfall tax and associated spending.
Source: HM Treasury, Public Finances Databank, January 2003, London, 2003 (www.hmtreasury.gov.uk/media//4EC32/jan03web.xls).
But how are we to judge whether the current fiscal position is consistent with
the golden rule, when we are in the midst of an uncompleted cycle? The
November 2002 Pre-Budget Report (PBR) measures progress ‘by the average
surplus on the current budget since 1999–2000, which on the government’s
provisional judgement is the start of the current cycle’.5 Over the first three
years of the cycle, the current budget averaged surpluses of 1.7% of national
income a year. The Treasury predicts deficits of 0.5% this year and 0.4% in
2003–04, reducing the average over the cycle to date to 0.8% next year. The
4
Whether the mini-cycle between quarter 1 1997 and mid-1999 survives future revisions must
remain somewhat in doubt, as output fell only fractionally below trend even at its weakest
point. As the Treasury concedes, it may turn out that the current cycle began in 1997 and that
the initial upswing was longer than current estimates suggest.
5
Paragraph 2.51 of HM Treasury, Pre-Budget Report: 2002, Cm. 5664, London, 2002
(www.hm-treasury.gov.uk/Pre_Budget_Report/prebud_pbr02/prebud_pbr02_index.cfm).
9
Green Budget, January 2003
average since 1999–2000 is then expected to stabilise at 0.7% until the cycle
ends.
This implies that the golden rule will be overachieved comfortably during the
current cycle, even if the upswing is weaker than expected and the negative
output gap takes a year or two longer to close. In cash terms, the Chancellor
said in November that he expected to end the current cycle with a cumulative
current budget surplus of £46 billion.6 This implies he could run cumulative
current deficits of nearly £50 billion between now and the end of the cycle in
2005–06 without breaking the golden rule, rather than the £2.7 billion he
currently projects.
But assessing compliance with the golden rule in this way can be unhelpfully
backward-looking. Should the government really determine its fiscal room for
manoeuvre by asking what deficits it can ‘afford’, given the size of the
surpluses it has accumulated in the recent past? And what happens when the
current cycle ends? Does the Treasury start again with a blank sheet of paper
or does it argue that the deficits accumulated in the second half of the current
cycle have to be offset by surpluses in the first half of the next?
Economic cycles are an endlessly repeating process and any point in time
could, in principle, be taken as the start of a new cycle or the end of an old
one. In reaching their Budget judgements, chancellors confront the legacy of
past fiscal policy decisions and external shocks in the form of the public debt
and its servicing costs. But provided the debt is not in itself on an
unsustainable path (for example, as long as the debt-to-GDP ratio is stable
over time, which is a requirement of the sustainable investment rule), it is not
clear that the current balances recorded in the past should dictate the current
balances permitted in the future. More relevant is whether the current budget
is on course to be in balance or surplus on average looking forward – or
whether tax increases and/or spending cuts are needed to ensure that it is.
The International Monetary Fund made a similar point in the concluding
statement of its Article IV review of the UK economy in December: ‘It would
be useful to assess whether the rules could be designed, or calibrated, so as to
reduce their dependence on past over-performance, avoiding the risk that
margins accumulated in the past allow excessive leeway in the future’.7
Symmetrically, it could have added that the rules should not mean that past
deficits necessarily constrain the government to run an undesirably tight fiscal
policy unless there is a risk of missing the debt-to-GDP target.
Whether fiscal policy is consistent with the golden rule looking forward is best
assessed by focusing on estimates of the current budget balance that are
adjusted (albeit imperfectly) for the impact of the economic cycle. The
Treasury estimates that a 1% fall in national income increases public sector net
borrowing (the current budget balance plus net investment) by 0.5% of
6
Chancellor of the Exchequer’s Pre-Budget Report Statement, 27 November 2002 (www.hmtreasury.gov.uk/Pre_Budget_Report/prebud_pbr02/prebud_pbr02_speech.cfm).
7
International Monetary Fund, ‘United Kingdom – 2002 Article IV Consultation Concluding
Statement’, 9 December 2002 (www.imf.org/external/np/ms/2002/120902a.htm).
10
Planning the public finances
national income after one year and 0.2% after two years.8 So, given an
estimate of the output gap, the Treasury can estimate what the underlying
budget position would be if national income were at its trend level.
The cyclically adjusted current balance is also shown in Figure 2.2. It suggests
that the current budget moved into surplus in the late 1980s only because of
the extent to which the economy was operating above trend, which inflated tax
revenues and cut social security costs. Similarly, it suggests that the big rise in
the deficit in the early 1990s reflected both the impact of the recession and a
weaker underlying fiscal position. Over the present economic cycle, the actual
current balance and the cyclically adjusted current balance are projected to
have moved, and to continue to move, pretty closely in line. This reflects
lower volatility in national income, and therefore smaller output gaps.
The Treasury points to its forecast of an average current budget surplus of
0.7% of national income over this cycle as evidence of adequate caution in the
setting of the public finances. The same could be said of its projections for the
cyclically adjusted current budget, which is still just in surplus this year and
next and which rises to 0.7% of national income by 2007–08. Projections of
cyclically adjusted surpluses in future years could be interpreted as a
reasonable indication that current policy is consistent with the golden rule,
whatever the pattern of deficits or surpluses in the past.
This, of course, requires that those projections are realistic and that they
contain an appropriate margin for error, given the uncertainties inherent in
determining the size of the output gap and in distinguishing between cyclical
and underlying movements in the public finances. In the Treasury’s view, the
golden rule will still be met even if the trend level of output turns out to be 1%
lower than currently projected and if past surpluses therefore owed more to
strong economic activity than to the underlying health of the public finances.
The average current balance since the start of the cycle would remain positive
throughout. Less reassuringly, the cyclically adjusted current balance would
be consistently in deficit and only move back to balance by 2007–08.
Whether 0.7% of national income is an adequate cushion is a matter of
judgement. It would not have been large enough to avoid missing the golden
rule given some errors in estimating trend output in the past. But the reduced
volatility of national income in the latest cycle suggests such errors are
perhaps unlikely to be as large as they were in the late 1980s and early 1990s.
Let us assume, for the sake of argument, that trend output has been estimated
correctly. Just as important a question in judging whether policy is now
consistent with the golden rule is whether it is realistic to expect the cyclically
adjusted current balance to rise to 0.7% over the next few years, as forecast in
the PBR, without further tax increases or cuts in spending plans. (If this level
were achieved and then maintained, the government would have the same
level of caution built into its projections looking forward as it has signalled it
was looking for in the past.) This prediction may be unduly reliant on
8
Paragraph B17 on page 185 of HM Treasury, Pre-Budget Report: 2002, Cm 5664, London,
2002 (www.hmtreasury.gov.uk/Pre_Budget_Report/prebud_pbr02/prebud_pbr02_index.cfm).
11
Green Budget, January 2003
ambitious forecasts – for example, for corporation tax receipts and movements
in equity and housing markets, which are discussed in Section 2.2.
Assessing whether the sustainable investment rule is met
While the golden rule imposes a binding constraint on the government, this is
not currently true of the sustainable investment rule. Figure 2.3 shows public
sector net debt as a share of national income from 1974–75 to 2007–08. This
has declined steadily from 43.7% of national income in March 1997, just
before Labour took office, to 30.4% of national income in March 2002. Net
debt is set to rise over the next six years, to reach 33.0% of national income in
March 2008 according to Treasury plans for receipts and spending. Assuming
that the golden rule is met, the government would need to spend an additional
7.0% of national income on public sector net investment to breach the
sustainable investment rule. Given the difficulty that the government is having
in increasing public sector net investment spending from an historical low of
0.5% of national income in 1999–2000 to 2.0%, such an increase seems highly
unlikely.
Figure 2.3. Public sector net debt as a percentage of national income
80
Percentage of GDP
70
60
50
40
30
20
10
0
74–75 77–78 80–81 83–84 86–87 89–90 92–93 95–96 98–99 01–02 04–05 07–08
Financial year
Source: HM Treasury, Public Finances Databank, January 2003, London, 2003 (www.hmtreasury.gov.uk/media//4EC32/jan03web.xls).
Errors made in the past in predicting budget balances
In the 2002 PBR last November, the Treasury revised down the estimates of
the current budget balance for this year and next that it published in the
Budget last April. For the current financial year, it now forecasts a current
deficit of 0.5% of national income, rather than a surplus of 0.3%, and for
2003–04, it forecasts a 0.4% deficit, rather than a 0.6% surplus. Both revisions
are almost entirely due to lower-than-expected revenues. These forecasts are
still consistent with cyclically adjusted surpluses in each year, but of only
0.2% and 0.3% of national income respectively. It would take only a small
12
Planning the public finances
forecasting error to see a cyclically adjusted deficit in one or both years. If
such a deficit were maintained indefinitely, the golden rule would be missed.
Table 2.1 shows the Treasury’s average error in forecasting public sector net
borrowing (PSNB) – the current balance plus net investment – one, two, three
and four years ahead between 1985–86 and 1997–98. It shows that even one
year ahead, the average absolute error is £12.5 billion in today’s prices. Even
when the effect of misforecasting national income has been stripped out, there
is still an average error of £10.4 billion. Underestimating borrowing by even
half that magnitude for 2003–04 would lead to a cyclically adjusted deficit on
current budget.9
Table 2.1. Average errors in forecasting public sector net borrowing, as a
percentage of national income and in £ billion
Time period
Average error
(% GDP)
Average error
(£bn)
Average error,
Average error,
correct GDP
correct GDP
(% GDP)
(£bn)
One year ahead
1.2
12.5
1.0
10.4
Two years ahead
2.0
20.9
1.4
14.6
Three years ahead
3.0
31.3
2.0
20.9
Four years ahead
4.1
42.8
2.4
25.1
Notes: Figures in £ billion are calculated assuming HM Treasury GDP forecast for 2002–03 of
£1,044 billion. Average error corresponds to the average absolute error over the period 1985–
86 to 1997–98.
Source: Table B13 of HM Treasury, Pre-Budget Report, November 1998, Cm. 4076
(http://archive.treasury.gov.uk/pub/html/prebudgetNov98/index.html).
For the last two fiscal years, 2000–01 and 2001–02, the error in forecasting
PSNB has been slightly lower, averaging 0.9% of national income. The Endof-Year Fiscal Report, published by HM Treasury alongside the November
2002 Pre-Budget Report, shows the errors made in forecasting borrowing one
year ahead from 1989–90 to 2001–02. Over these years, the average error was
1.1% – slightly lower than the 1.2% presented in Table 2.1. Not surprisingly,
borrowing tends to be underestimated when economic activity is weak.
The Private Finance Initiative and the interpretation of the
fiscal rules
The fiscal rules constrain policy by requiring that public sector deficits and
debt do not exceed particular levels over the economic cycle. Yet the
definition of public sector net debt used does not measure the full cost of
activities carried out on behalf of the public sector. This raises the possibility
that the government could circumvent the fiscal rules by financing spending
through liabilities that are not covered by the sustainable investment rule. This
is possible under the Private Finance Initiative (PFI).
If a public investment project is paid for by conventional deficit financing,
then the cost will appear immediately as public sector investment, and add to
both public sector net borrowing and public sector net debt. In future years,
9
If the error were in public sector net investment, then this would not lead to a change in the
current budget surplus.
13
Green Budget, January 2003
the depreciation of the asset purchased, and the additional debt interest
repayments, will score as current expenditure and reduce the size of any
current budget surplus.
Alternatively, the government might use the private sector to finance and
deliver the project. In this scenario, the cost of the capital spending would
initially be paid for by the PFI contractor rather than scoring against public
sector net borrowing. This would reduce public sector net borrowing and net
debt compared with what they would have been under the conventional
finance route. Initially, this would make the sustainable investment rule easier
to meet.
Use of the PFI would also reduce depreciation costs, as the public sector
capital stock is lower. The reduction in public sector net debt would also
reduce future debt interest payments. But the annual stream of payments to the
PFI contractor would score as current expenditure. If a private contractor were
able to deliver a public sector investment project with efficiency gains that
precisely offset their higher borrowing costs, then the project would cost
exactly the same as under conventional finance. Hence it would not make the
golden rule easier or harder to meet.10 The only economic rationale for using
the PFI is that it is hoped it will offer better value for money. Despite the fact
that the private sector faces higher borrowing costs, these are expected to be
offset by its greater operating efficiency.11 If this holds, the private contractor
can provide a given standard of public service at a lower cost, which will, all
other things remaining equal, lead to lower levels of public borrowing.
Table 2.2. Forecast capital spending by the public sector under both
conventional finance and the Private Finance Initiative
2002–03 2003–04 2004–05 2005–06
Public sector net investment (£bn)
14.3
19.6
21.9
24.1
Depreciation (£bn)
14.1
14.7
15.4
16.2
Asset sales (£bn)
3.8
3.8
3.8
3.8
Public sector gross investment (£bn)
32.2
38.1
41.1
44.1
Capital spending by the private sector (£bn)
3.7
3.3
2.5
2.5
Total publicly sponsored gross investment (£bn)
35.9
41.4
43.6
46.6
PFI investment as a % of total publicly sponsored
10.3%
8.0%
5.7%
5.4%
gross investment
Notes: Figures on capital spending by the private sector correspond to signed deals only.
Figure for 2005–06 is an estimate.
Source: Table 2.1 on page 10 of HM Treasury, Spending Review 2002: Departmental
Investment Strategies: A Summary, Cm. 5674, Stationery Office, London, 2002 (www.hmtreasury.gov.uk/media//343A6/dis_whitepaper02.pdf).
Most publicly sponsored investment is still conducted through conventional
means rather than through the PFI. In 2002–03, total publicly sponsored
10
Assuming that payments are structured so that the annual payments to the private contractor
correspond to the flow of services received.
11
For a discussion of issues arising from private sector involvement in the delivery of public
services, see chapter 3 of A. Dilnot, C. Emmerson and H. Simpson (eds), The IFS Green
Budget: January 2002, Commentary no. 87, Institute for Fiscal Studies, London, 2002
(www.ifs.org.uk/gb2002/chap3.pdf).
14
Planning the public finances
investment is expected by the Treasury to be £35.9 billion, of which just
£3.7 billion, or 10.3%, will be financed by the private sector. Table 2.2 shows
the share of public investment financed through the PFI declining. This is
because the figures are based on deals that have been signed so far. Whether
the share actually declines will depend on how many new PFI contracts are
agreed. If current policy continues, we can expect to see a steady increase in
the number of PFI contracts signed, leading to higher forecasts for government
spending on the PFI.
On the basis of deals signed so far, public sector net debt at the end of March
2006 would be 3.8% of national income higher if all the capital spending from
PFI deals signed so far been carried out using conventional finance.12 This is
shown in Figure 2.4. Should further PFI deals be signed, then net debt would
be higher still. But it would remain comfortably below the government’s 40%
ceiling unless there were an unfeasibly large increase in investment spending
(either financed conventionally or through the PFI) or unless the golden rule
were also breached, in which case some of the additional borrowing would
represent current rather than investment spending.
Figure 2.4. Public sector net debt as a percentage of national income
60
Public sector net debt plus capital spending carried out under the PFI
Public sector net debt
Percentage of GDP
50
40
30
20
10
0
90–91
92–93
94–95
96–97
98–99
00–01
02–03
04–05
Financial year
Source: HM Treasury, Public Finances Databank, January 2003, London, 2003 (www.hmtreasury.gov.uk/media//4EC32/jan03web.xls). Capital spending carried out under the PFI from
T. Clark, M. Elsby and S. Love, ‘Trends in public investment’, Fiscal Studies, 2002, vol. 23,
pp. 305–42.
So far, we have focused on what would have happened to the public debt had
conventional finance been used to deliver all of the public sector investment
projects that have been financed through PFI deals. An alternative approach is
to look at the stream of payments that the government is committed to paying
PFI providers in return for the services that they are providing.
12
This assumes that the public sector could have carried out the investment spending at the
same cost as the private provider.
15
Green Budget, January 2003
The expected payments over the next 25 years from PFI contracts that have
already been agreed are shown in Figure 2.5. It shows that if no further
contracts are agreed, payments should rise to 0.45% of national income in
2003–04 before falling to less than 0.1% of national income in the mid-2020s.
Again, as with the figures on capital spending by the private sector, this
decline is unlikely to materialise because new contracts should be signed. We
can therefore expect upward revisions to the projected future flow of payments
under the PFI as we have seen in recent Budgets, as is shown in Figure 2.5.
Figure 2.5. Estimated future payments under Private Finance Initiative
contracts as a percentage of national income
0.50
Percentage of national income
2002 Budget
2001 Budget
0.40
2000 Budget
0.30
0.20
0.10
0.00
1998–99
2003–04
2008–09
2013–14
Year
2018–19
2023–24
Sources: Table C15 on page 208 of HM Treasury, Financial Statement and Budget Report,
HC346, March 2000; Table C18 on page 206 of HM Treasury, Financial Statement and
Budget Report, HC279, March 2001; Table C10 on page 232 of HM Treasury, Financial
Statement and Budget Report, HC592, April 2002 (www.hmtreasury.gov.uk/budget/bud_index.cfm).
Summing the payments due until March 2028 (as a share of national income)
gives a total of 5.9% of national income, equivalent to £62 billion in today’s
prices. But only around 22% of these payments are to finance the initial capital
investment.13 The rest will pay for current goods and services – for example,
the delivery of usable hospital beds rather than a hospital building.
Do the government’s contractual obligations to fund this spending mean that
the £62 billion (5.9% of national income) should be regarded as a component
of public sector net debt? If so, it would bring the debt level far closer to the
ceiling of 40% of national income set by the Chancellor’s sustainable
investment rule. It should be noted, though, that £26 billion out of this
£62 billion represents payments that are due to be made before March 2008.
These are included in the government’s published plans for future current
13
Arthur Anderson and Enterprise LSE, Value for Money Drivers in the Private Finance
Initiative, London, 2000.
16
Planning the public finances
spending and therefore score against both future public sector net borrowing
and future public sector net debt.
Including the estimated stream of future payments would provide an estimate
of the increase in public sector net debt that we might expect to see if the
government decided to pay for all of these contracts upfront and continue to
receive the services that they provide in the future.14 But it is far from clear
that including them in public sector net debt would be particularly sensible.
For one thing, there are large parts of public spending that the government
(and all political parties) are, in practice, no less committed to than payments
to PFI contractors. Examples of such spending include the provision of many
public services (such as some healthcare) free at the point of use and increases
in many welfare benefits (such as the basic state pension) at least in line with
inflation. These de facto liabilities are not added to public sector net debt –
and if they were, they would add up to a very large proportion of national
income. They are paid for conventionally out of future national income when
the services they provide are consumed.15
One could also argue that future payments to PFI providers are different from
these other liabilities because they are contractual obligations. But if the
government were to decide that it no longer wanted the services that it had
bought from these private providers, it should be able to negotiate a substantial
discount on the contracted payments because they would no longer have to
provide them. The provider would doubtless still demand sufficient payment
from the government to cover the cost of the initial capital investment. In any
event, the future stream of payments due to PFI providers does not necessarily
represent an unavoidable cost to a future government that might not want to
consume some packages of services that its predecessors have signed up to.
On balance, including the stream of future payments as public sector net debt
on transparency grounds does not seem particularly sensible. It would also
provide an incentive for governments not to use the PFI even if this offered
demonstrably better value for money than conventional finance.
There is also nothing sacrosanct about the ceiling of 40% of national income
placed by the Chancellor on a particular measure of government debt.
Economic theory has little to tell us about what an optimal debt-to-GDP ratio
is, just that it should not be allowed to increase continuously over time. The
Chancellor could just as easily have chosen a different measure of debt, or a
different percentage. Indeed, if the Chancellor had chosen a measure of debt
that included the future financing of PFI contracts, he may well have chosen a
target higher than 40%.
14
The actual value would depend on the discount rate of the private sector. Using the
£62 billion calculated above would assume a discount rate equal to nominal growth in national
income of 4.8% (which comprises long-term real growth in national income of 2¼% and
inflation of 2½%).
15
Whether or not current government policy is sustainable in the medium and long term can
be examined, under a number of assumptions, using generational accounts. See HM Treasury,
Long-Term Public Finance Report: An Analysis of Fiscal Sustainability, November 2002
(www.hmtreasury.gov.uk/Pre_Budget_Report/prebud_pbr02/assoc_docs/prebud_pbr02_adsustain.cfm).
17
Green Budget, January 2003
The Stability and Growth Pact
Should the UK decide to join the Euro, fiscal policy might be further
constrained by the Stability and Growth Pact. This requires member countries
to set a medium-term budgetary objective of ‘close to balance or surplus’. This
implies either higher taxes or lower public spending than required by the
golden rule since it prohibits the government from borrowing to invest.
Joining the Euro also formally requires the UK government to comply with the
Excessive Deficits Procedure in the Maastricht Treaty. This requires that
general government gross debt should be below 60% of national income and
that deficits should be below 3% of national income. Should these criteria not
be met (either in terms of actual outcomes or in terms of plans), the European
Council could choose to apply penalties such as fines.
The danger with a balanced-budget rule is that it could inappropriately prevent
spending on beneficial investment projects that are prohibitively expensive for
current taxpayers alone to finance, because it would not permit future
generations to bear part of the cost.16 This seems unreasonable, given that
individuals often choose to fund long-term purchases by borrowing – for
example, paying for a house with a mortgage.
As we discussed earlier, the golden rule does make a distinction between
capital and current spending and therefore would allow such investment
projects to go ahead. This is not to say that the golden rule is necessarily ideal.
For one thing, it is hard to decide whether some categories of spending (e.g.
education) should be current or capital. Neither is it clear that the golden rule
delivers true intergenerational fairness, because the timing of the flow of
services financed by borrowing may not coincide with the timing of payments
on the debt.
It is possible that the balanced-budget rule – or at least its interpretation –
might be relaxed by the time the UK joined the Euro. EU President Romano
Prodi has stated that ‘the stability pact is stupid, like all decisions that are
rigid’.17 Also, in a November 2002 Communication to the European Council
and Parliament, the European Commission argued that the balanced-budget
rule should be interpreted using a cyclically adjusted measure of the budget
position and that countries with debt ‘well below the 60% reference level’
would be allowed small deviations from the balanced-budget rule. In parallel,
the Commission argued that the balanced-budget rule ‘could be required to
cater for the inter-temporal budgetary impact of large structural reforms (such
as productive investment or tax reforms) that raise employment or growth
potential’.18 While these changes would not make the Stability and Growth
16
An argument made in favour of a balanced-budget rule is that a golden rule would simply
provide an incentive for countries to redefine how they classify investment spending and, in
particular, how depreciation is measured. This is discussed in, for example, M. Buti, S.
Eijffinger and D. Franco, ‘Revisiting the Stability and Growth Pact: grand design or internal
adjustment?’, Centre for Economic Policy Research, Discussion Paper no. 3692, 2002
(www.cepr.org/pubs/new-dps/dplist.asp?dpno=3692).
17
Interview to the Le Monde newspaper, 17 October 2002.
18
Commission of the European Communities, Communication from the Commission to the
Council and the European Parliament, COM (2002) 668 final, Brussels, 27 November 2002.
18
Planning the public finances
Pact criteria identical to the UK fiscal rules, they would certainly bring them
closer – not least by potentially allowing higher deficits to finance investment
spending, provided debt is significantly lower than the 60% mark.
Table 2.3 shows how in 2001 the UK compared with other EU countries in
terms of both government borrowing and debt. In that year at least, the UK
had a relatively strong public balance (net borrowing or lending), with a
surplus of 0.8% of national income compared with a weighted average of –
0.8% across the entire EU and –1.5% across the Eurozone. Equally, when the
cyclically adjusted measure of the public balance is considered, the UK’s
surplus of 0.6% of national income compares favourably with both the EU
weighted average of –1.0% and the –1.5% average in the Eurozone countries.
The UK also had a relatively low level of general government gross debt
(39.1% of national income compared with a weighted EU average of 63.1%
and a weighted Eurozone average of 69.2%).
Table 2.3. Public finances across the EU in 2001, percentage of national
income
Public balance
Cyclically adjusted
Debt
public balance
Portugal
–4.1
–4.6
55.5
Germany
–2.8
–2.4
59.5
Italy
–2.2
–2.1
109.8
France
–1.4
–1.7
57.3
Greece
–1.2
–1.7
107.0
Spain
–0.1
–0.2
57.1
Netherlands
0.1
–0.2
52.8
Austria
0.2
0.1
63.2
Belgium
0.4
0.3
107.6
UK
0.8
0.6
39.1
Ireland
1.5
–0.1
36.4
Denmark
3.1
2.9
44.7
Sweden
4.8
4.7
56.6
Finland
4.9
5.1a
43.4
Luxembourg
6.1
n/a
5.6
Unweighted EU average
0.7
0.1b
59.7
Weighted EU average
–0.8
–1.0
63.1
b
Unweighted Eurozone
0.1
–0.7
62.9
Weighted Eurozone
–1.5
–1.5
69.2
a
Given as a proportion of mainland potential GDP. The figure excludes revenues from oil
production.
b
These unweighted averages exclude Luxembourg.
Note: Public balance refers to net borrowing or lending of consolidated general government
sector. Debt refers to general government consolidated gross debt.
Source: EUROSTAT website (http://europa.eu.int/comm/eurostat/) and OECD, Economic
Outlook, no. 72, December 2002 (www.oecd.org/EN/document/0,,EN-document-0nodirectorate-no-2-21578-0,00.html).
Even if the interpretation of the balanced-budget rule were to be relaxed – for
example, by allowing countries with a relatively low level of debt to borrow
more – the UK might still be constrained by the Excessive Deficits Procedure.
UK public sector net investment is forecast to grow to 2.0% of national
19
Green Budget, January 2003
income in 2005–0619 (which is the last year of the current Spending Review
period). Even with a balance on the cyclically adjusted current budget, the
limit on borrowing of 3% of national income would be being breached if
economic growth in 2004–05 were more than 1¼ percentage points lower than
expected and there were no bounce-back in the following year.20
2.2 Planning and forecasting revenues
Most tax revenues tend to follow the ups and downs of the economic cycle,
because they are levied on elements of incomes and spending that move
roughly in line with activity in the economy as a whole. For example, when
the economy is strong, more people will be in paid employment and paying
income tax and National Insurance. As earnings generally rise in excess of
inflation, over time people will also tend to be pulled into higher income tax
brackets. The number of higher-rate taxpayers has increased from 1.7 million
in 1990–91 to an estimated 3.1 million in 2002–03.21 This process of ‘fiscal
drag’ results from the progressive structure of the tax system and means that
income tax receipts tend to rise as a proportion of national income.
As we discussed in Section 2.1, it is possible to adjust revenue (and spending)
figures for the impact of the economic cycle and thereby better gauge the
underlying strength of the public finances. But this is never easy, and it may
be particularly difficult now. One important reason is that some tax revenues
are affected significantly by movements in asset markets, such as the housing
market and the stock market. These tend to be prone to longer, more
pronounced and less predictable cycles than economic activity as a whole.
This is evident from the boom and recent decline in the stock market, with
fears mounting in recent months that the housing market is due for a similar
setback after a lengthy period of strong growth.
A recent study published by the European Central Bank concluded that asset
price movements are an important determinant of budget balances in many
countries – over and above the impact of the cycle in national income.22 But
the study found that this effect was particularly marked in the UK. There is a
danger that the Treasury has underestimated the impact of asset price
movements on the public finances. This may mean that revenues bounce back
19
Table B6 on page 191 of HM Treasury, Pre-Budget Report: 2002, Cm. 5664, London, 2002
(www.hm-treasury.gov.uk/Pre_Budget_Report/prebud_pbr02/prebud_pbr02_index.cfm).
20
The Treasury estimates that a 1 percentage point reduction in growth increases borrowing
by 0.5% of national income in the first year and by a further 0.3% of national income the
following year. Therefore a 1¼ percentage point reduction in growth would increase
borrowing by 1% of national income in the following year. This, added to the 2% planned
investment spending, would lead to public sector borrowing of 3% of national income.
Source: Paragraph B17 of HM Treasury, Pre-Budget Report: 2002, Cm. 5664, London, 2002
(www.hm-treasury.gov.uk/Pre_Budget_Report/prebud_pbr02/prebud_pbr02_index.cfm).
21
Inland Revenue, Inland Revenue Statistics
(www.inlandrevenue.gov.uk/stats/income_tax/it_t01_1.htm).
22
F. Eschenbach and L. Schuknecht, ‘Asset prices and fiscal balances’, European Central
Bank, Working Paper no. 141, 2002 (www.ecb.int/pub/wp/ecbwp141.pdf).
20
Planning the public finances
less sharply than the Chancellor expects as the economy moves back to trend
output. This would make it harder to be confident of meeting the golden rule.
The stock market has a direct impact on revenues via stamp duty, which is
levied on share transactions (and also property transactions). We discuss both
types of stamp duty in greater detail below. The stock market also has an
indirect effect by influencing the fortunes of financial companies, which pay
corporation tax and whose employees contribute to income tax and National
Insurance receipts. In the November 2002 PBR, the government revised down
its revenue forecasts by £5 billion specifically because of the plight of
financial sector companies. This comes on top of a £3½ billion downward
revision for other stock market effects.
Although income tax revenues generally track movements in the economy, in
recent years they have come in more strongly than anticipated by the Treasury.
For example, the March 2001 Budget revised upwards its forecast of income
tax revenues in 2001–02 by £6 billion compared with the forecast made a year
earlier.23 If a significant proportion of the unexpected increases in income tax
receipts was due to the strength of the stock market, then there has to be some
concern about the path that income tax receipts will take in the future.
Corporation tax (CT) revenues have become increasingly reliant on the
financial sector too in recent years, with the share coming from that quarter
doubling from 18% to 36% between 1990 and 2000 alone.24 This helps
explain the recent weakness of CT revenues (which we discuss in greater
detail below) and suggests that the future path of CT revenues will depend
much more than in the past on the performance of the financial sector. Unless
one is willing to assume that stock markets and financial companies’ profits
soon return to the unusual levels of performance seen in the late 1990s, it
seems doubtful that corporation tax revenues will rebound as quickly as the
Chancellor expects.
Even if the stock market does recover strongly, boosting the financial sector,
there remain risks to the UK public finances from the uncertain prospects for
house prices. Even a relatively small fall in house prices, if associated with a
sharp decrease in the volume of sales, could have a big impact on stamp duty
revenues and a smaller impact on inheritance tax and capital gains tax.
Moreover, any downturn in the housing market could lead to a decline in
consumer confidence and a corresponding fall in revenues from VAT.
A third asset price effect could operate through the currency market. Some
analysts expect a sharp drop in sterling, accompanied by a rebalancing in the
economy away from consumer spending and towards exports. As the
government receives more tax revenue from each pound of the former than of
the latter, this too could have a sustained depressing effect on tax revenues.
We now turn to stamp duty and corporation tax in more detail.
23
See paragraph C37 of HM Treasury, Financial Statement and Budget Report, HC279,
March 2001.
24
Tables 11.4 and 11.5 of Inland Revenue, Inland Revenue Statistics, various years.
21
Green Budget, January 2003
Stamp duty
We can see the impact of asset markets on the UK public finances in the path
of revenues from stamp duty on property and share transactions. The tax bases
for these duties are directly related to the level of prices and volume of
transactions in the housing market and stock market respectively.
Figure 2.6. Revenues from stamp duty on shares and property
Percentage of national income
0.5
0.4
Stamp duties on stocks and shares
Stamp duties on property
0.3
0.2
0.1
0.0
1980-81 1983-84 1986-87 1989-90 1992-93 1995-96 1998-99 2001-02
Financial year
Source: Inland Revenue Statistics, various years.
Figure 2.7. Nominal growth in national income, the FTSE All-Share Index
and house prices
40.0
National income
FTSE All-Share Index
Halifax house price index
Nominal growth (%)
30.0
20.0
10.0
0.0
-10.0
-20.0
1980-81 1983-84 1986-87 1989-90 1992-93 1995-96 1998-99 2001-02
Financial year
Source: GDP – ONS, www.statistics.gov.uk; Halifax house price index – HBOSplc,
www.hbosplc.com/view/housepriceindex/housepriceindex.asp; FTSE – Thomson Financial
Datastream, www.datastream.com.
22
Planning the public finances
Revenues from these taxes between 1980–81 and 2001–02 are shown as a
share of national income in Figure 2.6. Annual growth in national income,
house prices and the FTSE All-Share Index are shown in Figure 2.7. As the
increases in national income, the FTSE All-Share and house prices are shown
in nominal terms, they will be higher at times of high inflation, other things
being equal.
The figures confirm that stamp duty revenues do not have a consistent
relationship with economic growth. Instead, stamp duty on property tracks
movements in house prices, although stamp duty on stocks and shares has, at
best, a weak lagged correlation with share prices. In the case of stocks and
shares, the volume of transactions plays an important role, and a high volume
need not be linked to a bull market. It should be noted that stamp duty on
shares was cut from 2% to 0.5% during the mid-1980s, while the stamp duty
rate on property has been increased four times since Labour came into power
in May 1997.25 But these changes do not alter the underlying picture.
For most of the years from 1980–81 to the late 1990s, stamp duty revenues
from stocks and shares were broadly constant as a proportion of national
income – despite the cuts in the rate of stamp duty on shares in the mid-1980s.
The peak observed in 1988–89 did coincide with strong economic growth,
both in the current and the previous financial year. But the dramatic revenue
increases observed since 1995–96 come at a time of unspectacular economic
growth. These increases in stamp duty revenues do, then, appear to be
explained by a rising stock market and also an increase in the number of
shares and share turnover, rather than by economic growth.26
Revenues from stamp duty on property transactions also diverge from the path
of the wider economy. From 1970–71 to the mid-1980s, they rose slowly.
From 1985–86 to 1995–96, they followed the economy with a lag of a year or
two. Since then, the pattern has been broadly in line with revenues from stamp
duty on stocks and shares, rising steadily despite average economic
performance. Despite the obvious contribution of increases in the rate of stamp
duty on property since May 1997, an important factor has been the rise in the
housing market, which has increased in value by over 75%. Taking out the
effect of the increases in rates would still leave revenues from property by
2001–02 being more than double the revenue in 1996–97.
Corporation tax
Corporation tax accounts for a relatively modest 12% of total tax revenues.27
But it plays an important role in explaining movements in the public finances
25
See Chapter 9 for more details on changes to stamp duty on property in recent years.
26
For more details and an assessment of the economic impact of stamp duty on shares, see M.
Hawkins and J. McCrae, Stamp Duty on Share Transactions: Is There a Case for Change?,
Commentary no. 89, IFS, London, 2002 (www.ifs.org.uk/corptax/comm89.pdf).
27
This is the average over calendar years 1987 to 2001. Data source: National Statistics,
Financial Statistics, The Stationery Office, London
23
Green Budget, January 2003
because it is one of the most volatile sources of tax revenue. More than half
the cut in revenue forecasts made in the 2002 PBR came in corporation tax.28
Whether or not such a fall is worrying will depend on whether corporation tax
revenues can be relied upon to bounce back soon. The government believes
they can. In 2001–02, corporation tax raised 2.9% of national income.29 The
Treasury expects this to fall to 2.5% for two years and then return to 2.9% in
2004–05. In the following two years, the forecasts predict even higher
revenues, of 3.2% and 3.3% of national income respectively. This sounds
reassuring, but is it credible? Since Budget 2001, corporation tax forecasts
have been cut on three consecutive occasions in the light of disappointing outturns. Each downward revision of the forecast for the near future has been
accompanied by an upward revision of the growth rate of revenues for later
years. This implies that almost the entire revenue decline is attributed to
cyclical or other temporary factors that will soon reverse.
But even if one were to accept that the entire fall in revenues were due to such
factors, the long-run forecast of 3.3% would still seem optimistic, as Figure
2.8 illustrates. This figure shows corporation tax revenues as a share of
national income over the last 15 years (thick black line) and predictions for the
following five years (thick grey line). Over the last 15 years, corporation tax
revenues averaged 3.2% of national income. While this seems very close to
the forecast, it ignores the fact that tax rates have been cut. Revenues also
received a temporary boost in the four years from 1999–2000 to 2002–03 from
the introduction of a payment system in quarterly instalments (see Chapter 9).
This had the effect of bringing tax payments forward and thus led to
companies paying more than one year’s tax per year during the transition to
the new system.
The dotted line shows an approximation of the revenues that would have been
raised with current tax rates and having stripped out the temporary effect of
the introduction of quarterly payments. These adjustments shift the graph
downwards in each of the last 15 years. The average yield, represented by a
thin grey line, drops to only around 2.9% of national income. This casts some
doubt on the long-term forecast of 3.3%. Not only is the forecast clearly above
the average of what the current system would have raised (despite the fact that
the Treasury does not currently forecast output rising above trend after 2005–
06), but also it is set so high that the forecast level would only have been
reached three times during the last 15 years according to our adjusted series.
The predicted level of corporation tax revenues for the long term therefore
seems unduly optimistic, especially given the importance of the financial
sector we noted earlier.
28
Forecasted tax revenues fell by £6 billion; the cut in forecasted corporation tax revenues
contributed £3.7 billion to this figure. Source: Table B9 of HM Treasury, Pre-Budget Report:
2002, Cm. 5664, London, 2002 (www.hmtreasury.gov.uk/Pre_Budget_Report/prebud_pbr02/prebud_pbr02_index.cfm).
29
This and the following figures exclude corporation tax revenues from the continental shelf
and are gross of enhanced and payable tax credits. Source: HM Treasury, Pre-Budget Report:
2002, Cm. 5664, London, 2002 (www.hmtreasury.gov.uk/Pre_Budget_Report/prebud_pbr02/prebud_pbr02_index.cfm).
24
Planning the public finances
Figure 2.8. Non-North-Sea corporation tax revenues and forecasts as a
share of national income
Percentage of national income
5.0%
Estimated outcomes
Adjusted series
Forecasts
Average (of adjusted)
4.0%
3.0%
2.0%
1.0%
0.0%
87–88 89–90 91–92 93–94 95–96 97–98 99–00 01–02 03–04 05–06 07–08
Year
Notes: The adjusted series shows the tax revenue that the current tax system would have
raised in previous years. It takes account of the fact that tax rates have been reduced in two
steps from 33% to 30% and that the introduction of a quarterly payments system boosted
revenues temporarily between 1999–2000 and 2002–03.
Sources: Inland Revenue, Inland Revenue Statistics, London, 2002; HM Treasury, Pre-Budget
Report: 2002, Cm. 5664, London, 2002 (www.hmtreasury.gov.uk/Pre_Budget_Report/prebud_pbr02/prebud_pbr02_index.cfm); Inland
Revenue, ‘A Modern System for Corporation Tax Payments’, Press Release IR 9, 17 March
1998.
Corporation tax revenues are very volatile and difficult to predict, and frequent
revisions are not surprising. It makes sense not to change long-term forecasts
every time short-term revenues change. But when revisions always go in the
same direction, the long-term forecasts need to be questioned at some point.
Looking further ahead, other forces may affect the UK’s ability to raise
revenue from corporate income taxes. These include more intense tax
competition between governments and possible action by the European Court
of Justice in response to company complaints or new European Union
directives.
2.3 Planning government spending
Since taking office in 1997, the government has substantially reformed the
planning and presentation of government spending, including the creation of
new aggregates with which to measure it. Total managed expenditure (TME)
is the broadest measure of public sector spending. At 40.2% of national
income in 2002–03, it is currently lower than the 41.0% spent in the last year
25
Green Budget, January 2003
of the Conservative government (1996–97). But TME is now set to rise to
42.3% by 2007–08.30
TME is divided into annually managed expenditure (AME) and departmental
expenditure limits (DELs), with around 45% of TME falling in AME and the
remaining 55% in DELs. AME is planned on an annual basis, as its name
suggests, and includes spending on items that are deemed more difficult to
plan years in advance, such as social security and debt interest payments.
DELs include spending plans for most government departments and, since
1998, are set up to three years in advance in biennial Spending Reviews. The
third Spending Review, in July 2002, revised the plans set out in July 2000 for
spending in 2003–04, and set out new spending limits for the years 2004–05
and 2005–06. The plans for 2005–06 are expected to be revised in the next
Spending Review, in July 2004, which will also unveil plans for 2006–07 and
2007–08. Unusually, we do have some information about spending in those
two years, as the April 2002 Budget set out the NHS spending plans for all
years until 2007–08. These allow for an average annual real increase in NHS
spending of 7.3% by 2007–08 compared with last year.
The point of having Spending Reviews that set DELs up to three years in
advance is that it should allow departments to plan their spending with a
longer-term perspective. In recent years, the government has added money to
DELs between Spending Reviews. But this is unlikely in the forthcoming
Budget due to the weakness of receipts, which has already led to upward
revisions in the borrowing forecasts, and due to the fact that the last Spending
Review was only in July 2002. Instead, we are likely only to see revisions to
AME.
Planning departmental expenditure limits
In the early years of the current planning regime, several departments failed to
spend their full allocations and overall spending under DELs came in lower
than planned. It was argued that this may have been partly due to the
introduction of End-Year Flexibility, which gave departments an entitlement
to spend any of their unspent allocation for a given year in subsequent years.
The idea was to remove the incentive that previously existed for departments
to spend all their money at the end of a financial year, even if they could have
obtained better value for money by spending it later.
Spending out-turn figures for 2001–02 show that underspending on current
items has ceased to be a problem. Figures to date for 2002–03 show that, if
anything, current spending is increasing faster than the plans allow.31
But the delivery of investment spending remains problematic. Figure 2.9
shows investment spending as a percentage of national income from 1979–80
to 2005–06 according to Treasury forecasts. Net investment as a percentage of
national income was volatile over most of the 1980s. After reaching just over
30
Table B6 on page 191 of HM Treasury, Pre-Budget Report: 2002, Cm. 5664, London, 2002
(www.hm-treasury.gov.uk/Pre_Budget_Report/prebud_pbr02/prebud_pbr02_index.cfm).
31
For more details, see IFS Public Finances Bulletin, 21 January 2003
(www.ifs.org.uk/press/pub_fin.shtml).
26
Planning the public finances
2% of national income in 1992–93, it fell to less than 0.6% by 1997–98. The
July 1998 Comprehensive Spending Review (CSR) included plans to increase
investment spending steadily over the next three years to nearly 1.1% of
national income by 2001–02. This is also shown in Figure 2.9. Despite these
plans, investment spending fell in the first year of the plans (1999–2000) to the
lowest level since 1988–89. Although investment spending increased in 2000–
01 and 2001–02, it was only in 2001–02 that spending was higher than when
Labour came into power in 1997. In 2001–02, net investment stood at
£8.8 billion – 30% lower than the £13 billion announced in the July 1998
CSR. In the first nine months of 2002–03, we have seen net investment
increase by 11.3% over the same months last year. In order to attain the level
expected under current plans, the overall increase in investment spending
would have to be 53.9% for the whole year. In order to attain this and spend
the £14.3 billion outlined in the November 2002 PBR, investment spending
for the remaining three months would have to be 97.5% higher than the same
period last year.
Figure 2.9. Public sector net investment: out-turns and forecasts as a
percentage of national income
Actual
July 1998 Comprehensive Spending Review
November 2002 Pre-Budget Report
2.5
Percentage of GDP
2.0
1.5
1.0
0.5
0.0
79–80
82–83
85–86
88–89
91–92
94–95
97–98
00–01
03–04
Financial year
Source: Out-turn figures from HM Treasury, Public Finances Databank, January 2003,
London, 2003 (www.hm-treasury.gov.uk/media//4EC32/jan03web.xls); forecasts from HM
Treasury, Pre-Budget Report: 2002, Cm. 5664, London, 2002 (www.hmtreasury.gov.uk/Pre_Budget_Report/prebud_pbr02/prebud_pbr02_index.cfm); July 1998
Comprehensive Spending Review figures from HM Treasury, Pre-Budget Report, Cm. 4076,
London, November 1998, as the plans set out in the actual Comprehensive Spending Review
were set using a different accounting system.
Public sector pay
Over the three years from 2002–03 to 2005–06, resource departmental
expenditure limits are due to increase by 7.6% a year on average in nominal
27
Green Budget, January 2003
terms.32 This represents a 5.1% average annual increase in real terms. It is
from within these increases that the government will have to fund any pay
increases in the public sector. It is clear from the increases planned in overall
spending that there is considerable scope for increasing the amount the
government spends on wages in the coming years.
These significant increases for resource DELs, initially unveiled in the July
2002 Spending Review, were presented to enable an improvement in the
quality of public services. There are some parts of the public sector where
relatively high pay increases may be necessary to improve services by
recruiting more high-quality staff or by motivating and retaining existing staff.
While large increases in the pay of particular public sector workers will be
easily affordable within the existing spending plans, large pay increases across
the board will not be. Total remunerations to public sector workers are in the
region of £115 billion a year – so each additional 1% increase in pay will use
just over £1 billion of resources available to the government.33 The
government should certainly weigh up carefully the benefits it might expect
from increases in wages against those it could expect to derive from spending
the money differently.
Issues in planning annually managed expenditure
Given the proximity of the Spending Review in July 2002, which set
departmental expenditure limits through to March 2006, it is likely that any
revisions to public spending announced in the Budget will be changes to
annually managed expenditure rather than changes to DELs.
Changes to AME might occur for at least two reasons. First, the Chancellor, as
discussed in Chapter 4, might decide that he would like to announce further
increases in financial transfers to low-income families with children so that the
government can continue to progress towards its child poverty targets.
Secondly, the Chancellor might decide that he needs to increase the funds he
has available in case of unforeseen contingencies.
This contingency reserve is known as the AME margin, and recent years have
seen it used to pay for the costs of the BSE crisis and the foot-and-mouth
epidemic. Since the 1998 Comprehensive Spending Review, the Chancellor
has tended to ensure that the margin contains £1 billion for the following
financial year, £2 billion for two years hence and £3 billion for three years
hence. Table 2.4 shows the size of the AME margin in recent years. Each
Spending Review (in 1998, 2000 and 2002) and Budget (1999, 2000, 2001 and
2002) has left the AME margin at £1 billion, £2 billion and £3 billion in one,
32
Table B17 of HM Treasury, Pre-Budget Report: 2002, Cm. 5664, London, 2002 (www.hmtreasury.gov.uk/Pre_Budget_Report/prebud_pbr02/prebud_pbr02_index.cfm).
33
According to the National Accounts Blue Book of 2000, total wages and employers’ social
contributions of those in public corporations, central government and local government were
£104 billion. Assuming 5% nominal growth a year over the last two years would imply
spending in 2002 of around £115 billion.
28
Planning the public finances
two and three years’ time respectively. As shown in last year’s Green Budget,
these levels of reserve are low by historical standards.34
At the time of the Pre-Budget Report, the Chancellor has tended to adjust the
AME margin to offset changes in the remainder of AME. So, for example, in
November 1998 and November 1999, falls in forecast expenditure on AME
led to increases in the size of the margin. In the following Budgets, the
Chancellor was able to ‘reset’ the AME margin back to its normal level and
use the funds to cut taxes, increase public spending or reduce debt. At the time
of the 2002 Budget, £0.5 billion of additional funds were needed to restore the
AME margin in 2003–04 to £2 billion.
Table 2.4. The size of the AME margin (£ billion)
1999–
2000
1.0
3.0
1.0
3.5
0.0
2000–
01
2.0
4.5
2.0
3.9
1.0
2001–
02
3.0
6.0
3.0
6.4
2.0
1.0
2.7
1.0
0.2
0.0
0.0
2002–
03
2003–
04
2004–
05
Comprehensive Spending Review, July 1998
Pre-Budget Report, November 1998
Budget, March 1999
Pre-Budget Report, November 1999
Budget, March 2000
Spending Review, July 2000
2.0
3.0
Pre-Budget Report, November 2000
3.6
4.6
Budget, March 2001
2.0
3.0
Pre-Budget Report, November 2001
1.2
1.5
Budget, April 2002
1.0
2.0
Spending Review, July 2002
1.3
1.0
2.0
Pre-Budget Report, November 2002
0.1
1.8
0.5
Possible Budget 2003 scenario
1.0
2.0
Addition to spending?
–0.8
+1.5
Sources: Various HM Treasury Pre-Budget Report, Budget and Spending Review
documentation.
2005–
06
3.0
0.5
3.0
+2.5
Should the Chancellor wish to restore the AME margin back to its normal
levels in the 2003 Budget, then he will need to find an additional £1.5 billion
in 2004–05 and an additional £2.5 billion in 2005–06. This will require an
increase in borrowing, an increase in taxation or a reduction in spending
elsewhere of the same magnitude.
2.4 Conclusions
Under its current plans, the government is set to meet both the golden rule and
the sustainable investment rule over the current economic cycle. But the
golden rule is forecast to be met with less margin for error than in previous
years. The likelihood of deficits on the current budget in the near future
underlines the importance of judging whether policy at any given time is
consistent with the golden rule looking forward. The sustainable investment
rule poses less of a constraint – even if we were to add the future liabilities of
the government for capital spending undertaken under the Private Finance
34
See figure 2.7 in A. Dilnot, C. Emmerson and H. Simpson (eds.), The IFS Green Budget:
January 2002, Commentary no. 87, IFS, London (www.ifs.org.uk/gb2002/chap2.pdf).
29
Green Budget, January 2003
Initiative. The fiscal rules are less constraining than the Stability and Growth
Pact as it is currently interpreted, although this interpretation may already have
changed by the time the UK joins the Euro – if it ever does.
The government expects revenues to recover as the economy picks up. But
movements in asset markets complicate the task of assessing the underlying
health of tax revenues and the public finances more generally. The
Chancellor’s forecasts show stamp duty and corporation tax receipts
increasing. As a significant part of these revenues is linked to the performance
of the stock market and financial companies, it may be risky to rely too much
on this rebound. The public finances might well turn out to be less healthy if
the stock market does not deliver these missing revenues, while the possibility
of a decline in house prices poses a further risk. On top of any concern about
revenues, the Chancellor must also decide whether to allocate resources to
restore his contingency reserve to the levels he has felt necessary in the past.
Robert Chote, Carl Emmerson, Christine Frayne and Alexander Klemm
30
3. IFS public finance forecasts
In his Pre-Budget Report last November, the Chancellor conceded that the
impact of a weak economy on tax revenues would force him to borrow more
this year and next. But, over the medium term, he predicted that the economy
would bounce back – and that the public finances would bounce back with it.
This chapter details the January 2003 IFS public finance forecasts and
compares them with the Treasury’s November 2002 Pre-Budget Report (PBR)
projections. Taking the two fiscal rules and the degree of caution that the
Chancellor has previously chosen to incorporate into his plans, we consider
whether he will need to announce reductions in public spending or further tax
increases.
Our main conclusions are as follows:
1. In the short run, our forecasts are very similar to those in the PBR. In
2002–03, we forecast public sector net borrowing of £22.1 billion,
slightly higher than the PBR forecast of £20.1 billion. In 2003–04, we
forecast public sector net borrowing of £25.2 billion, compared with
£24.5 billion forecast in the PBR.
2. In the medium term, we believe that the public finances will be weaker
than the PBR suggested in November, even if the economy behaves
much as the Treasury expects. In 2005–06, we forecast public sector
net borrowing of £28 billion, compared with the PBR forecast of
£19 billion.
3. Nonetheless, we believe the Chancellor can credibly claim that both
the golden rule and the sustainable investment rule will have been met
comfortably over the current economic cycle, which is projected to end
in 2005–06.
4. If the Chancellor sticks by his PBR forecast of a surplus on current
budget in 2005–06 of 0.4% of national income, he could say that there
is no need for any further tax increases. But looking forward into the
next economic cycle, our forecasts imply that spending cuts or tax
increases will be required if the golden rule is to continue to be met.
5. Spending cuts would sit oddly with both the government’s stated
objectives and its previous actions. Tax increases, therefore, seem
more likely.
6. The scale of tax increases required depends on how cautious the
government wants to be. To expect to continue to meet the golden rule
exactly would require tax increases of around £4 billion. But in the
past, the Chancellor has been more cautious and sought to overachieve
the golden rule by around 0.7% of national income on average. To do
this would require tax increases of around £11 billion.
31
Green Budget, January 2003
3.1 Borrowing in 2002–03
In 2001–02, government revenues came in lower than forecast either by the
Treasury in the November 2001 PBR or by IFS in the January 2002 Green
Budget. Public spending also came in lower than the Treasury expected, but
slightly higher than forecast by IFS. The net effect was to leave public sector
net borrowing in balance, compared with the £1.4 billion deficit predicted by
the Treasury and the £1.6 billion surplus predicted by IFS.1
Receipts have continued to come in lower than expected during the current
financial year. In the PBR last November, the Treasury cut its April Budget
forecast for current receipts in 2002–03 by £7.5 billion to £399.7 billion, as
shown in Table 3.1. On the spending side, it revised its Budget forecast for
total managed expenditure up by £1.4 billion to £419.8 billion. This largely
reflected the carrying forward of last year’s underspending in departmental
expenditure limits, plus a £1 billion special allocation to help finance a
possible conflict in Iraq.
Table 3.1. Comparison of Green Budget and HM Treasury forecasts for
government borrowing, 2002–03 (£ billion)
Budget,
Apr. 02
PreBudget
Report,
Nov. 02
Green
Budget,
Jan. 03
Differences in Green
Budget forecast
relative to:
Budget
PBR
Current receipts
407.2
399.7
396.6
–10.6
–3.1
Total managed expenditure
418.4
419.8
418.7
0.3
–1.1
Of which:
Departmental expenditure limits
239.7b
241.3
240.3
0.6
–1.0
Annually managed expenditure
178.7b
178.5
178.4
–0.3
–0.1
Public sector net borrowinga
11.2
20.1
22.1
10.9
2.0
Net investment
14.4
14.3
13.3
–1.1
–1.0
Surplus on current budgeta
3.2
–5.7
–8.8
–12.0
–3.1
a
Includes windfall tax and associated spending.
b
DEL and AME forecast adjusted for accounting changes introduced between the Budget and
the Spending Review. For more details, see table B14 on page 202 of the 2002 Pre-Budget
Report.
Sources: Treasury forecasts – HM Treasury, Pre-Budget Report: 2002, Cm. 5664, London,
2002 (www.hmtreasury.gov.uk/Pre_Budget_Report/prebud_pbr02/prebud_pbr02_index.cfm), and HM
Treasury, Financial Statement and Budget Report, HC592, London, 2002 (www.hmtreasury.gov.uk/budget/bud_bud02/budget_report/bud_bud02_repindex.cfm).
Weaker receipts and higher spending imply higher borrowing. The Budget had
forecast public sector net borrowing of £11.2 billion and a surplus on the
current budget of £3.2 billion. The PBR showed public sector net borrowing of
£20.1 billion and a current budget deficit of £5.7 billion.
1
See Appendix A for a breakdown of the November 2001 Pre-Budget Report and the January
2002 IFS Green Budget forecast for 2001–02.
32
IFS public finance forecasts
Table 3.2. Comparison of Green Budget and HM Treasury forecasts for
government borrowing, 2002–03 and 2003–04 (£ billion)
2002–03
PBR
Gr. Budget
Nov. 2002
Jan. 2003
2003–04
PBR
Gr. Budget
Nov. 2002
Jan. 2003
Inland Revenue
Income tax (gross of tax credits)
114.1
110.0h
123.0
118.5h
a
h
Corporation tax (CT)
29.3
28.5
30.8
30.0h
Tax creditsb
–3.5
n/a
–4.9
n/a
Petroleum revenue tax
1.1
1.1
1.3
1.3
Capital gains tax
2.0
2.0
1.4
1.4
Inheritance tax
2.4
2.4
2.6
2.6
Stamp duties
8.2
7.8
8.6
8.2
Social security contributions
65.5
65.0
75.4
75.1
Total Inland Revenue (net of tax credits)
219.1
216.8
238.3
237.1
Customs and Excise
Value added tax (VAT)
64.5
63.7
67.3
67.0
Fuel duties
22.4
22.4
23.1
23.1
Tobacco duties
8.2
8.2
7.8
8.2
Spirit duties
2.2
2.2
2.4
2.4
Wine duties
1.9
1.9
1.9
1.9
Beer and cider duties
3.1
3.1
3.1
3.1
Betting and gaming duties
1.3
1.3
1.3
1.3
Air passenger duty
0.8
0.8
0.8
0.8
Insurance premium tax
2.1
2.1
2.2
2.2
Landfill tax
0.5
0.5
0.7
0.7
Climate change levy
0.9
0.9
0.9
0.9
Aggregates levy
0.2
0.2
0.4
0.4
Customs duties and levies
2.0
2.0
1.9
1.9
Total Customs and Excise
110.1
109.3
113.8
113.9
Vehicle excise duties
4.4
4.4
4.8
4.8
Oil royalties
0.5
0.5
0.0
0.0
Business ratesc
18.0
18.0
18.2
18.2
Council tax
16.6
16.6
17.8
17.8
Other taxes and royaltiesd
10.9
10.9
12.3
12.3
Total taxes and social security contribnse
379.6
376.5
405.1
404.1
Accruals adjustments on taxes
–0.6
–0.6
3.4
3.4
Less Own resources contribution to EU
–3.0
–3.0
–2.4
–2.4
Less Public corporations’ CT payments
–0.2
–0.2
–0.2
–0.2
Tax creditsf
1.2
1.2
0.6
0.6
Interest and dividends
4.1
4.1
4.1
4.1
Other receipts
18.6
18.6
19.7
19.7
Current receipts
399.7
396.6
430.3
429.3
Current spending
405.5
405.4
435.2
434.4
Current balanceg
–5.7
–8.8
–4.9
–5.1
Net investment
14.3
13.3
19.6
20.1
Public sector net borrowingg
20.1
22.1
24.5
25.2
a
National accounts measure: gross of enhanced and payable tax credits. b Includes enhanced and
payable company tax credits. c Includes district council rates in Northern Ireland. d Includes money paid
into the National Lottery Distribution Fund. e Includes VAT and ‘traditional own resources’
contributions to EC budget. Cash basis. f Excludes children’s tax credit and other tax credits that score
as a tax repayment in the National Accounts. g Includes expenditure associated with the windfall tax.
h
Net of tax credits.
Note: For more details of the IFS forecast in 2002–03, see Table A.3 in Appendix A.
Sources: Treasury forecasts from HM Treasury, Pre-Budget Report 2002, Cm. 5664, London, 2002
(www.hm-treasury.gov.uk/pre_budget_report/prebud_pbr02/prebud_pbr02_index.cfm) – this table is
similar to table B12 on page 197); authors’ calculations.
33
Green Budget, January 2003
IFS now forecasts receipts this year of £396.6 billion, a further £3.1 billion
below the Treasury’s PBR prediction. We also expect total managed
expenditure to come in £1.1 billion below the PBR forecast, at £418.7 billion.
This implies a deficit on the current budget of £8.8 billion and public sector
net borrowing of £22.1 billion, worse than the PBR forecast by £3.1 billion
and £2.0 billion respectively.
Table 3.2 compares our revenue forecasts for 2002–03 and 2003–04 with
those of the PBR. Ours are gloomier for this year because we expect lower
income tax, corporation tax, stamp duty and VAT receipts than projected in
the PBR.
We also have a lower forecast for public spending, dominated by a predicted
underspend of £1 billion on capital spending within departmental expenditure
limits. This reflects the fact that public sector net investment has proved
difficult to deliver during the year to date.2 We also assume that the £0.1
billion in the annually managed expenditure margin in 2002–03 at the time of
the November 2002 PBR will not be spent.
3.2 Borrowing in 2003–04
Wherever possible, our forecasts for 2003–04 are based on the same
macroeconomic assumptions that underlie the ‘cautious’ forecasts in the 2002
PBR, including the assumption that the trend rate of economic growth is 2½%.
We assume that no new measures are announced in the Spring 2003 Budget.
But, as discussed in Chapter 4, the government might decide to increase
further the generosity of payments to lower-income families with children to
increase its chance of hitting its child poverty target.
We forecast receipts of £429.3 billion in 2003–04, £1 billion lower than the
£430.3 billion in the PBR. This is a smaller undershoot than we forecast in
2002–03 because we expect stronger revenue growth than the Treasury for a
number of taxes, such as tobacco duties. But we do expect weaker receipts
from corporation tax and stamp duty than the PBR, as we are less confident
that the loss of revenue attributed by the Treasury to the performance of
financial companies will be recouped.
On the expenditure side, we forecast current spending of £434.4 billion next
year, £0.8 billion less than in the PBR. This is because we assume that the
AME margin for 2003–04 will be set to £1.0 billion, rather than the
£1.8 billion in the PBR. As shown in Table 2.4 in Chapter 2, this is line with
recent Budget practice. But we forecast that net investment spending will be
£0.5 billion higher than the PBR forecast, at £20.1 billion, because we assume
that half the £1 billion underspend we forecast for this year will be carried
forward and spent in 2003–04, with the remainder spent in 2004–05.
With our projections of lower receipts and lower spending than in the PBR
partially offsetting each other, our predictions for borrowing are little different
2
For more details, see IFS Public Finances Bulletin, 21 January 2003
(www.ifs.org.uk/press/pub_fin.shtml).
34
IFS public finance forecasts
from the Treasury’s November forecasts. For 2003–04, we expect a deficit on
the current budget of £5.1 billion (compared to the Treasury’s £4.9 billion)
and public sector net borrowing of £25.2 billion (compared to the Treasury’s
£24.5 billion).
3.3 Medium-term prospects
Our forecasts for the current budget balance and public sector net borrowing
this year and next are not very far out of line with those in the PBR. But
looking further ahead, we are rather less optimistic. As with our short-term
forecasts, our medium-term projections (shown in Table 3.3 later) are based
on similar macroeconomic assumptions to those the Treasury used in the PBR.
Further details can be found in Appendix A.
One of the key uncertainties in projecting the outlook for the public finances is
the path of corporation tax revenues, traditionally one of the hardest taxes to
forecast. As discussed in Section 2.2, in recent years the Treasury has forecast
strong growth in corporation tax receipts, taking them to a level that by
historical standards appears rather high.
In last year’s IFS Green Budget, we said, ‘Due to the difficulties in forecasting
corporation tax in the current environment, we take the November 2001 PBR
forecasts, although … it would be helpful if the Treasury published further
discussion of what is driving the forecast increase in revenues in the medium
term.’3 The Treasury has not published further details of its forecasts. In the
January 2003 Green Budget, we assume that the Treasury’s forecast growth in
corporate tax revenues will not materialise. Instead, we assume that, as the
economy returns to trend, corporate tax receipts will return to their average
level of recent years. This implies annual nominal growth in underlying
corporation tax receipts of around 8.5% a year to March 2008. This is lower
than the growth rate implied by recent Treasury forecasts.
Figure 3.1 shows non-North-Sea corporation tax receipts adjusted for changes
to the corporation tax system from 1987–88 to 2001–02, along with the
Treasury and IFS projections for 2002–03 to 2007–08. The recent average in
adjusted corporation tax receipts is 2.9% of national income. By 2007–08,
when the economy is still forecast to be at trend, the Treasury expects
corporation tax receipts of 3.3% of national income – a level only achieved in
the past at the peak of the economic cycle. We forecast corporation tax
receipts of just over 2.9% in 2007–08 (when the economy is assumed to be at
trend output). The difference between the two projections equals 0.3% of
national income, or £3.5 billion in today’s prices.
3
Page 25 of A. Dilnot, C. Emmerson and H. Simpson (eds), The IFS Green Budget: January
2002, Commentary no. 87, IFS, London, 2002 (www.ifs.org.uk/gbfiles/gb2002.shtml).
35
Green Budget, January 2003
Figure 3.1. Non-North-Sea corporation tax receipts and forecasts as a
percentage of national income
5.0%
Percentage of GDP
4.0%
3.0%
2.0%
1.0%
0.0%
87–88
89–90
91–92
93–94
IFS forecasts
95–96
97–98
99–00
Treasury forecasts
01–02
03–04
05–06
07–08
Adjusted series
Notes: The adjusted series shows our estimates of the tax revenue that the current tax system
would have raised in previous years. It takes account of the fact that tax rates have been
reduced in two steps from 33% to 30% and that the introduction of a quarterly payments
system boosted revenues temporarily between 1999–2000 and 2002–03.
Sources: Inland Revenue, Inland Revenue Statistics, London, 2002; HM Treasury, Pre-Budget
Report: 2002, Cm. 5664, London, 2002 (www.hmtreasury.gov.uk/Pre_Budget_Report/prebud_pbr02/prebud_pbr02_index.cfm); Inland
Revenue, ‘A Modern System for Corporation Tax Payments’, Press Release IR 9, 17 March
1998.
On VAT receipts, the Treasury has introduced a new strategy for combating
VAT fraud and avoidance. It expects that this will produce more than £2
billion a year in additional revenues by 2005–06. In order to be cautious, the
Treasury has only assumed that £1.4 billion extra is collected by 2005–06.4
Our forecasts take a more cautious approach and assume that none of this
revenue materialises. Offsetting this, at least to some extent, is the fact that our
forecasts assume that the ratio of VAT receipts to consumer spending remains
constant over time, whereas the Treasury forecasts assume a gradual decline.
Our medium-term forecasts, like those of the Treasury, are for VAT receipts to
be stable, at around 6.1% of national income.
Over the medium term, we expect receipts overall to be lower than the
Treasury forecast in the PBR, despite our very similar forecast for 2003–04
(see Table 3.3). By 2005–06, when the economy is predicted to return to trend,
we forecast that receipts will be £6.5 billion lower than forecast in the PBR, at
£486.5 billion. We expect a slightly lower shortfall of £6.1 billion in 2007–08,
when the current planning period for the public finances ends.
4
HM Treasury, Pre-Budget Report: 2002, Cm. 5664, London, 2002 (www.hmtreasury.gov.uk/Pre_Budget_Report/prebud_pbr02/prebud_pbr02_index.cfm).
36
IFS public finance forecasts
Table 3.3. Medium-term public finances forecasts, based on cautious
macroeconomic assumptions (£ billion)
IFS forecasts
Current budget
Current receipts
Current expenditurea
Surplus on current budgetb
Capital budget
Net investment
Public sector net borrowingb
2002–03
2003–04
2004–05
2005–06
2006–07
2007–08
396.6
405.4
–8.8
429.3
434.4
–5.1
459.1
461.5
–2.4
486.5
490.6
–4.0
514
518
–4
542
546
–5
13.3
22.1
20.1
25.2
22.4
24.8
24.1
28.1
27
31
30
35
HM Treasury forecasts
Current budget
Current receipts
399.7
430.3
463
493
521
548
Current expenditurea
405.5
435.2
459.7
487.5
513
538
Surplus on current budgetb
–5.7
–4.9
3
5
8
10
Capital budget
Net investment
14.3
19.6
21.9
24.1
27
30
Public sector net borrowingb
20.1
24.5
19
19
19
20
a
In line with the National Accounts, depreciation has been included as current expenditure.
b
Includes spending financed by the windfall tax.
Sources: Treasury forecasts from HM Treasury, Pre-Budget Report: 2002, Cm. 5664, London,
2002 (www.hm-treasury.gov.uk/Pre_Budget_Report/prebud_pbr02/prebud_pbr02_index.cfm)
– this table is similar to table B5 on page 190; authors’ calculations.
In forecasting spending, we take the Treasury’s current departmental
expenditure limits to 2005–06, with an additional £0.5 billion of capital
spending in both 2003–04 and 2004–05. We also assume that the AME margin
will be set to £2.0 billion in 2004–05 and £3.0 billion in 2005–06, as shown in
Table 2.4 of Chapter 2. This implies £1.5 billion of additional spending in
2004–05 and £2.5 billion of additional spending in 2005–06. We also take
account of any differences in debt interest spending arising from previous
differences in borrowing levels. The result is that, in 2005–06, we forecast
current spending of £490.6 billion, £3.1 billion higher than the PBR forecast.
Forecasting public spending in 2006–07 and 2007–08 is more difficult. It
requires an assumption about the level of funds that the Chancellor will wish
to allocate in the 2004 Spending Review. The PBR assumes that departmental
expenditure limits excluding expenditure on the NHS will fall as a share of
national income. With annually managed expenditure, the PBR assumes
annual growth of ‘1¾ per cent in real terms in line with its recent trend’,5
which is lower than expected GDP growth.
The PBR therefore assumes that non-NHS spending will fall as a share of
national income. It remains to be seen whether this is consistent with the
government’s stated objectives of reducing child poverty and delivering
‘world-class’ public services. The Chancellor might decide that more funds
are required to meet the government’s objectives, as he did in the April 2002
Budget. It seems appropriate that these decisions be left until the 2004 Budget,
5
Paragraph B20 on page 187 of HM Treasury, Pre-Budget Report: 2002, Cm. 5664, London,
2002 (www.hmtreasury.gov.uk/Pre_Budget_Report/prebud_pbr02/prebud_pbr02_index.cfm).
37
Green Budget, January 2003
by which point more information on how the government is progressing
towards its objectives will be available. In the mean time, we assume that all
non-NHS current spending grows in line with national income in 2006–07 and
2007–08. As shown in Table 3.3, this implies current spending of £546 billion
in 2007–08, some £8 billion above the PBR estimate.
With lower forecasts for receipts and higher forecasts for spending, we believe
that the public finances will be weaker in the medium term than the Treasury
currently projects. In 2005–06, we forecast a deficit on the current budget of
£4.0 billion rather than the surplus of £5 billion forecast by the Treasury. In
2007–08, we project a current deficit of £5 billion, while the Treasury has a
current surplus rising to £10 billion. We expect public sector net borrowing to
rise to £28.1 billion in 2005–06 and then to £35 billion in 2007–08, whereas
the Treasury has it rising only from £19 billion to £20 billion over the same
period.
Table 3.4. Medium-term public finances forecasts, based on cautious
macroeconomic assumptions (% of national income)
IFS forecasts
Current budget
Current receipts
Current expenditurea
Surplus on current budgetb
Average surplus on current
budget since 1999–2000b
Capital budget
Net investment
Public sector net borrowingb
Public sector net debt
Output gapc
2002-03
2003-04
2004-05
2005-06
2006-07
2007-08
38.0
38.8
–0.8
1.1
39.2
39.6
–0.5
0.8
39.6
39.9
–0.2
0.6
39.9
40.2
–0.3
0.5
40.1
40.4
–0.3
0.4
40.4
40.7
–0.3
0.3
1.3
2.1
31.2
–1.3
1.8
2.3
32.4
–1.0
1.9
2.1
33.2
–0.3
2.0
2.3
34.1
0.0
2.1
2.4
35.1
0.0
2.2
2.6
36.5
0.0
HM Treasury forecasts
Current budget
Current receipts
38.3
39.3
40.0
40.4
40.7
40.8
Current expenditurea
38.8
39.7
39.7
40.0
40.0
40.1
Surplus on current budgetb
–0.5
–0.4
0.3
0.4
0.6
0.7
Average surplus on current
1.2
0.8
0.6
0.7
0.7
0.7
budget since 1999–2000b
Capital budget
Net investment
1.4
1.8
1.9
2.0
2.1
2.2
Public sector net borrowingb
1.9
2.2
1.6
1.6
1.5
1.5
Public sector net debt
31.0
32.1
32.4
32.6
32.7
33.0
Output gapc
–1.3
–1.0
–0.3
0.0
0.0
0.0
a
In line with the National Accounts, depreciation has been included as current expenditure.
b
Includes spending financed by the windfall tax.
c
Measured as a percentage of trend output rather than actual output.
Sources: Treasury forecasts from HM Treasury, Pre-Budget Report: 2002, Cm. 5664, London,
2002 (www.hm-treasury.gov.uk/Pre_Budget_Report/prebud_pbr02/prebud_pbr02_index.cfm)
– this table is similar to table B5 on page 190; authors’ calculations.
Table 3.4 shows the IFS and Treasury forecasts as percentages of national
income. The higher borrowing forecast by IFS translates into debt being 0.2
percentage points higher in 2002–03 than the 31.0% of national income that
the Treasury expects. By 2007–08, the IFS forecast is for debt to stand at
38
IFS public finance forecasts
36.5% of national income, which is 3.5 percentage points higher than the
33.0% forecast by the Treasury. Nevertheless, under both the Treasury’s and
IFS’s forecasts, net debt is set to stay below the 40% level stipulated by the
sustainable investment rule.
Our forecasts suggest that there will have been an average surplus on the
current budget of 0.5% of national income between 1999–2000 and 2005–06 –
the Treasury’s estimate of the current economic cycle – as shown in Table 3.4.
In cash terms, the cumulative surplus of £31 billion is less than the £46 billion
the Chancellor predicted in the PBR, but over the current economic cycle, the
golden rule is still comfortably overachieved. However, we see the present
cycle ending in 2005–06 with the current budget in deficit to the tune of 0.3%
of national income, compared with the surplus of 0.4% and rising expected by
the Treasury. Our projections therefore imply that further tax increases or
spending cuts will be needed to ensure that the golden rule is not missed
looking forward.
3.4 The Budget judgement
The current budget balances forecast in the April 2002 Budget, the November
2002 Pre-Budget Report and the January 2003 IFS Green Budget are shown in
Figure 3.2. The PBR predicts that the golden rule will continue to be met, but
with less room to spare than looked likely at the time of the April 2002
Budget. We expect the current budget to remain in deficit. The large current
budget surpluses seen in 1999–2000, 2000–01 and 2001–02 would still allow
the government to claim that the golden rule had been met over the current
economic cycle. But our forecast implies that the current budget will remain in
deficit from 2005–06 onwards during a period when the economy is operating
at trend. Without tax increases or spending cuts at some point, the golden rule
will be missed.
So if the government has to raise taxes or cut spending, which is it likely to
choose? Cutting public spending significantly would seem inconsistent with
the government’s stated objectives, as it would involve big cuts in either
transfer payments or spending on public services.
Within transfer payments, the largest recipient groups are families with
children and pensioners. Meeting the government’s target for reducing child
poverty by one-quarter by 2004–05 might require further increases in spending
(see Chapter 4). The government has also pledged increases in support to
those aged 65 or over, including the introduction of the pension credit in
October 2003. Reductions in other transfer payments – for example, disability
or unemployment benefits – are possible but would need to reduce recipients’
incomes significantly to make substantial savings.
Looking at public services, the recent July 2002 Spending Review announced
supposedly fixed spending plans covering all departmental expenditure limits
to March 2006, apart from the NHS, where the settlement runs until March
2008. Changing these plans so soon would sit very uncomfortably with the
government’s commitments to sustained increases in public spending in order
to deliver world-class public services.
39
Green Budget, January 2003
Figure 3.2. Current budget surplus as a percentage of national income
2.0%
Treasury April 2002 Budget forecast
Current balance as a % of GDP
Treasury November 2002 Pre-Budget Report forecast
IFS January 2003 forecast
1.0%
0.0%
-1.0%
-2.0%
2001–02
2002–03
2003–04
2004–05
2005–06
2006–07
2007–08
Year
Sources: Treasury forecasts from HM Treasury, Pre-Budget Report: 2002, Cm. 5664, London,
2002 (www.hmtreasury.gov.uk/Pre_Budget_Report/prebud_pbr02/prebud_pbr02_index.cfm), and HM
Treasury, Financial Statement and Budget Report, HC592, London, 2002 (www.hmtreasury.gov.uk/budget/bud_bud02/budget_report/bud_bud02_repindex.cfm); authors’
calculations.
Tax increases therefore seem more likely, particularly given the choices the
Chancellor has made in the past: the July 1997 Budget increased taxes in order
to reduce borrowing, while the April 2002 Budget increased taxes in order to
increase public spending.
How big do these tax increases need to be? Let us assume that the Chancellor
wants to end the current economic cycle in 2005–06 in a position that would
not require him to make any further tax increases or spending cuts to hit the
golden rule over the following economic cycle. This would imply a current
budget balance in that year. If he sticks by his PBR forecast of a surplus on
current budget in 2005–06 of 0.4% of national income, the Chancellor could
say that there is no need for any further tax increases. But, on our forecasts, he
would need to close the current deficit of 0.3% of national income that we
expect for that year. This would require a tax increase of about £4 billion,
effective by April 2005.
But in the past, the Chancellor has been more cautious than this. He has sought
to build caution into his projections by aiming for a current surplus of around
0.7% of national income on average, sufficient to ensure that the golden rule is
met even if the Treasury has overestimated the level of trend output by 1%.
On the basis of his PBR forecast, this could be seen as requiring a tax increase
of around £3 billion to raise the projected current surplus in 2005–06 from
0.4% of national income to 0.7%. But the Chancellor might also point out that
he expects the surplus to rise to 0.7% of national income in any event by
40
IFS public finance forecasts
2007–08, even without the economy moving above trend. So again there is no
need for tax increases.
But on our forecast, the Chancellor would need to raise taxes by around
£11 billion to move from a current deficit of 0.3% of national income to a
current surplus of 0.7% – and thereby restore the level of caution he has
sought in the past with regard to the golden rule. We do not expect the
underlying budget position to improve of its own accord beyond 2005–06, in
part because we are not as confident as the Treasury that the revenues lost
from the stock market and the troubles of the financial sector will bounce
back. In addition, we expect spending to grow more quickly than the economy
in the 2004 Spending Review – although the Chancellor will not have to make
a firm decision on this until next year. A rise in spending as a share of national
income would offset fiscal drag, the natural tendency for revenues to rise as a
share of national income as people move into higher tax brackets.
To conclude, if the Chancellor believes that the forecasts he made in the PBR
are still realistic – and that the recent weakening of the public finances is a
temporary phenomenon – he could argue that there is no need for significant
tax increases. He might well add that with the outlook for the world economy
fragile, now is not the time for a tightening of fiscal policy that could put too
much of the burden of supporting the UK economy on interest rates.
But we believe that the outlook for the public finances is weaker than the PBR
suggested. If the golden rule is expected to be met going forwards, then taxes
would need to rise. Now may not be the ideal time for this, but rises cannot be
delayed for too long without undermining the credibility of those very rules in
which the Chancellor places such store.
Robert Chote, Carl Emmerson and Christine Frayne
41
4. What do the child poverty targets
mean for the child tax credit?
The child tax credit will be introduced in April 2003. When fully operational,
it will represent the majority of government financial support for children. It is
designed to help reduce child poverty and is means-tested against family
income.
The Labour government has an explicit target for the level of child poverty –
defined in terms of relative incomes – for 2004–05. This chapter therefore
examines whether the government’s current commitments are likely to enable
it to meet this target, and what it might cost to reduce child poverty further by
increasing the rates of the child tax credit or through other means in and
beyond April 2004. Our calculations suggest that around £1 billion of further
spending might be needed to meet the target – a sizeable amount of money
given the current state of the public finances, but small compared with the
total increases in spending on benefits for families with children since 1997.
Yet, regardless of whether the government is able to find more for families
with children, child poverty in 2004–05 should still be substantially lower than
it was in 1996–97.
4.1 Child poverty under Labour
In his Beveridge speech at Toynbee Hall in March 1999, the Prime Minister
announced a radical ambition – to ‘eradicate child poverty within a
generation’. Subsequently, the Treasury set out further objectives: to eradicate
child poverty by 2020, to halve it by 2010 and to ‘make substantial progress
towards eliminating child poverty by reducing the number of children in
poverty by at least a quarter by 2004’.1
The government has yet to specify how we might tell whether poverty has
been eradicated in 2020 and what measure of poverty is due to be halved by
2010, although the Department for Work and Pensions has been consulting on
a new measure of poverty and is due to announce its conclusions shortly.2 But
the government’s target for 2004–05 is more specific: ‘the target for 2004 will
be monitored by reference to the number of children in low-income
1
From HM Treasury, Spending Review 2000: Public Service Agreements 2001–04, Cm. 4808,
Stationery Office, London, 2000 (www.hmtreasury.gov.uk/Documents/Public_Spending_and_Services/Public_Service_Agreements_200
1-2004/pss_psa_whitepaper.cfm). The target was initially a joint Public Service Agreement
target for HM Treasury and the former Department of Social Security accompanying the 2000
Spending Review, and was carried forward in the 2002 Spending Review. In this, and the rest
of the chapter unless otherwise specified, ‘child’ means ‘dependent child’ – a child under 16,
or under 19 and in full-time education.
2
Department for Work and Pensions, Measuring Child Poverty: A Consultation Document,
London, 2002 (www.dwp.gov.uk/consultations/consult/2002/childpov/childpoverty.pdf).
42
Child poverty targets and the child tax credit
households by 2004/5. Low-income households are defined as households
with income below 60% of the median, as reported in the Households Below
Average Income (HBAI) statistics … Progress will be measured against the
1998/9 baseline figures and methodology’.3 The data to assess these targets
should be available in early 2006. In the HBAI statistics, children are
considered poor depending on the total income of the household in which they
live, leaving aside the question of whether money channelled to low-income
families with children is actually beneficial to the children themselves.
The wording of the government’s target does not specify whether income is to
be measured before housing costs (BHC) or after housing costs have been
deducted (AHC), nor whether self-employed households are to be included. In
practice, ministers’ statements have tended to focus on progress on the AHC
measure including the self-employed.4 As there were 4.2 million children in
poverty in 1998–99 on this definition, there will need to be fewer than
3.1 million children in poverty in 2004–05 to meet the target.5
What has happened to date? Between 1996–97 and 2000–01 (the latest year
for which data are currently available), the number of children in households
below 60% of median AHC income fell from 4.4 million to 3.9 million.6 This
implies that if the government is to reach its target in 2004–05 (AHC), child
poverty will have to continue to fall each year by an average of around
200,000 – faster than in the period from 1996–97 to 2000–01. Measuring
poverty BHC, the target level is 2.3 million children, and the government is
closer to reaching its target for 2004–05, having seen a decline from
3.1 million in 1998–99 to 2.7 million in 2000–01.7 As Figure 4.1 illustrates,
3
HM Treasury, ‘Technical Note for HM Treasury’s Public Service Agreement 2003–2006’,
London, 2002, www.hm-treasury.gov.uk/mediastore/otherfiles/tech_notes.pdf. The median
household is the one for which half the rest of the population has an income higher than it
does and half has an income lower.
4
For example, page 87, box 5.3 of HM Treasury, Budget Report 2001: Investing for the Long
Term, London, 2001 (www.hmtreasury.gov.uk/Budget/Budget_2001/Budget_Report/bud_bud01_repchap5.cfm?).
5
The data in this chapter on children in poverty up to 2000–01 come from M. Brewer, T.
Clark and A. Goodman, The Government’s Child Poverty Target: How Much Progress Has
Been Made?, Commentary no. 88, IFS, London, 2002
(www.ifs.org.uk/inequality/childpov.pdf), or M. Brewer, T. Clark and A. Goodman, ‘What
really happened to child poverty in the UK in Labour’s first term?’, Economic Journal,
forthcoming. Some of the numbers can also be found in Department for Work and Pensions,
Households Below Average Income 1994/95 to 2000/01, Corporate Document Services,
Leeds, 2002 (www.dss.gov.uk/asd/hbai/hbai2001/hbai2000_01.html).
6
Tests reveal that the fall in child poverty since 1996–97 is statistically significant – in other
words, it is unlikely to be driven by random variations in the data used to analyse poverty
rates.
7
Poverty is lower when measured BHC than when measured AHC because housing costs are
a larger proportion of total expenditure of low-income families with children. Because the
government and commentators tend to focus on measuring incomes AHC when thinking about
the bottom end of the income distribution, we do so in this chapter, reporting the results BHC
only where they are substantially different.
43
Green Budget, January 2003
the fall since 1996–97 follows a very long period during which child poverty
grew substantially.8
Figure 4.1. Child poverty (children in households with less than 60%
median income AHC)
Children (million)
5.0
4.0
3.0
2.0
1.0
0.0
1964
1968
1972
1976
1980
1984
Child poverty
1988
1992
1996
2000
Target for 2004-05
Notes: The poverty line of 60% median income AHC is a fraction of the contemporary median
household income across the whole population (i.e. not just for children). Data up to and
including 1992 are for calendar years, while those thereafter are for financial years, so 1993
should be read as 1993–94, and so on. Changes in the income definitions make comparisons
before and after 1993–94 difficult.
Source: Authors’ calculations based on Family Expenditure Survey for years until 1993 and
on Family Resources Survey thereafter.
4.2 The child tax credit
The structure of new tax credits in April 2003
In Budget 2002, the government confirmed that it will be introducing two new
tax credits from April 2003: the child tax credit and the working tax credit.
The child tax credit will bring together three parts of the existing tax and
benefit system that support families with children. Separately, the working tax
credit will support adults with or without children in low-paid work, as well as
providing subsidies for certain childcare expenditure for some working parents
(see Box 4.1).
8
See also P. Gregg, S. Harkness and S. Machin, ‘Poor kids: trends in child poverty in Britain,
1968–96’, Fiscal Studies, 1999, vol. 20, pp. 163–87.
44
Child poverty targets and the child tax credit
Box 4.1. The new tax credits from April 2003
• What is being abolished?
The children’s tax credit reduces the income tax bills of around 5 million
income-tax-paying families with children under 16. It was introduced in April
2001 and will be subsumed within the child tax credit in April 2003.
The working families’ tax credit (WFTC) provides support to 1.34 million
low-paid families with dependent children working 16 or more hours a week.
It will be subsumed within the child tax credit and the working tax credit in
April 2003.
The childcare tax credit subsidises some eligible childcare costs of 167,000
families who also receive the WFTC. It was introduced in October 1999 as
part of the WFTC and will be subsumed within the working tax credit in April
2003.
Child allowances in income support provide extra money to 1.18 million
families with dependent children claiming income support and income-related
jobseeker’s allowance. The child allowances will be subsumed within the
child tax credit in April 2004.
• What is new?
The child tax credit will provide income-related support to the main carer. By
April 2004, around 5.75 million – all but the richest 10% – of families with
dependent children should receive it. The working tax credit will provide inwork support to around 1 million single people or couples in low-paid work.
Both will be introduced in April 2003, although families on income support
will not be affected by the reforms until April 2004.
• What is staying?
Child benefit is a universal, non-means-tested payment. All 7 million families
with dependent children in the UK are entitled to receive it. It will be
unaffected by the reforms.
Note: A dependent child is one under 16, or under 19 and in full-time education.
Sources: Inland Revenue, Working Families Tax Credit Statistics Quarterly Enquiry, May
2002, London, 2002 (www.inlandrevenue.gov.uk/wftctables/wftc_may_02.pdf); Department
for Work and Pensions, Income Support Quarterly Statistical Enquiry, August 2002, London,
2002 (www.dwp.gov.uk/asd/asd1/qse/aug2002/is_aug2002_pub.pdf); HM Treasury, The
Child and Working Tax Credits, The Modernisation of Britain’s Tax and Benefit System no.
10, London, 2002 (www.hm-treasury.gov.uk/mediastore/otherfiles/new_tax_credits.pdf).
Budget 2002 announced that, in April 2003, the child tax credit will consist of
two components:9
9
This chapter does not discuss how these tax credits work in detail; for that, see HM Treasury,
The Child and Working Tax Credits, The Modernisation of Britain’s Tax and Benefit System
no. 10, London, 2002 (www.hm-treasury.gov.uk/mediastore/otherfiles/new_tax_credits.pdf),
or D. Thurley, ‘Tax credits and income changes’, Welfare Rights Bulletin, no. 169, Child
Poverty Action Group, August 2002. Childcare subsidies are also discussed in Chapter 7. M.
Brewer, T. Clark and M. Myck, Credit Where It’s Due? An Assessment of the New Tax
Credits, Commentary no. 86, IFS, London, 2001 (www.ifs.org.uk/taxben/taxcred.pdf), and
45
Green Budget, January 2003
•
a family element of £10.45 per week, doubled in the financial year of a
child’s birth;
•
an amount per dependent child of £27.75 per week.
Families with incomes below £13,230 p.a. (£254.42 p.w.) are entitled to the
full amount. Incomes above £13,230 p.a. reduce entitlement to the per-child
elements at the rate of 37p in the pound until a family is left with just the
family element (this happens at £17,130 p.a. for a one-child family and at
£21,030 p.a. for a two-child family).10 Incomes above £50,000 p.a. (£961.54
p.w.) reduce entitlement to the family element at the rate of 6.7p in the pound,
meaning that families with children with incomes over £58,110 (or £66,221 if
they have a child under 1) will be entitled to child benefit only.
Around 90% of families with children will be entitled to some child tax credit:
around half of families will be entitled to the per-child element and the family
element, and around 40% will be entitled to the family element only. The
structures of the existing system and of the new tax credits are illustrated in
Figures 4.2 and 4.3.
Figure 4.2. Financial support for a family with one child under the
outgoing system (£ per week)
Benefit and tax credit income
£120
£100
£80
£60
£40
WFTC IS adult
C
WFTC child
F
Children's tax credit
£20
Child benefit
£0
£0
£200
£400
£600
Gross income
£800
£1,000
£1,200
Notes: IS is adult allowance, F is the family premium and C is the child allowance in income
support. The figure uses hypothetical April 2003 rates for the WFTC and the children’s tax
credit. It does not show childcare tax credit.
chapter 5 of A. Dilnot, C. Emmerson and H. Simpson (eds), The IFS Green Budget: January
2002, Commentary no. 87, IFS, London, 2002 (www.ifs.org.uk/gb2002/chap5.pdf), (both
written before the government confirmed the precise structure of the new tax credits) analysed
the likely impact of the new tax credits on work incentives, incomes and take-up.
10
These three annual thresholds are higher if a family qualifies for extra credits for working at
least 30 hours a week, having disabled adults or children, or receiving help for childcare costs.
46
Child poverty targets and the child tax credit
Figure 4.3. Financial support for a family with one child under the new
tax credits (£ per week)
Benefit and tax credit income
£120
£100
£80
£60
Working
IS tax credit
Child tax
credit
£40
£20
Child benefit
£0
£0
£200
£400
£600
£800
£1,000
£1,200
Gross income
Notes: The new tax credits are annual systems but this figure assumes them to be weekly. It
also assumes that the new tax credits are implemented fully in April 2003 and that a family
qualifies for working tax credit at 16 hours of minimum-wage work (£67.20 p.w.) and for the
30-hour premium at £126 p.w. It does not show support for childcare costs. It assumes one
dependent child aged at least 12 months.
The future structure and rates
The structure of the child tax credit effectively means that the government has
two targeted instruments with which to redistribute to families with children
(as well as child benefit). The family element can be used to direct money to
the vast majority of families in a near-universal way, regardless of income or
family size. The child element, on the other hand, is focused on low- to
middle-income families, and also helps larger families more than smaller ones.
This suggests that, for a given level of expenditure, increasing the per-child
element will have a larger direct impact on poverty than increasing the family
element (we show this in Section 4.4). Perhaps for this reason, the government
has promised to increase the per-child element (£27.75 p.w.) in line with
average earnings growth until the end of this Parliament. This means that the
rate from April 2004 will be around 63p higher in real terms than the rate in
2003 if earnings rise in line with their trend of recent years.11 This is a much
smaller annual increase in child-related payments than families saw between
11
The average annual growth in nominal earnings between 1998 and 2001 is 4.74%. This
would imply a per-child element of £29.05 in April 2004. The inflation rate averaged over the
same period is 2.43%, implying that the real rise is some 63p.
47
Green Budget, January 2003
1996 and 2003, and equates to a real increase in incomes of only 0.7% for the
poorest couple with two children (0.8% for a lone parent with two children).12
4.3 Is the government likely to meet its
child poverty target in 2004–05?
The government has chosen to target a relative measure of child poverty rather
than an absolute one. Whether a household is judged to be poor depends not
only on the cash value of its income, but also on how much it has relative to
the median household. This means we cannot judge whether the government is
going to meet the target simply by looking at the impact of its own policy
reforms on the incomes of poor families with children. We also need to assess
the likely impact of economic and demographic factors on the incomes of poor
families and on the median income against which they are to be compared.
This distinction is sometimes overlooked. For example, in Budget 2001, the
Treasury claimed that reforms introduced from 1997 to 2001 would reduce the
number of children in poverty by 1.2 million. This number was widely
believed to be a forecast of how much child poverty would fall. But the
Treasury subsequently had to clarify that it was only an estimate of how much
higher child poverty would have been if the government had only increased
benefits and tax credits in line with inflation.13 In other words, it abstracted
from the impact of economic and other demographic factors on poverty rates.
Changes in these other factors are important for two reasons. First, some
changes – such as increased employment amongst mothers – will increase the
incomes of low-income families with children and directly reduce child
poverty.14 Secondly, changes in median income can have a large impact on
child poverty by ‘moving the goalposts’. Figure 4.1 shows that child poverty
fell by only 500,000 between 1996–97 and 2000–01, based on a relative
poverty line. If the government had instead chosen to fix the poverty line at its
1996–97 level in cash terms, then it could have claimed that child poverty had
fallen by 1.4 million.15
Unfortunately, accurately forecasting what might happen to the distribution of
income by 2004–05 is very difficult. Median household income is affected by
12
Total income defined as income support, child benefit and the child tax credit.
13
See HM Treasury, Tackling Child Poverty: Giving Every Child the Best Possible Start in
Life, London, 2001 (www.hmtreasury.gov.uk/mediastore/otherfiles/TacklingChildPoverty.pdf).
14
D. Piachaud and H. Sutherland, ‘Changing poverty post-1997’, CASEpaper 63, London
School of Economics, Centre for Analysis of Social Exclusion, 2002, show that changes in
employment patterns amongst families with children were very important in reducing child
poverty between 1996 and 2000.
15
See M. Brewer, T. Clark and A. Goodman, The Government’s Child Poverty Target: How
Much Progress Has Been Made?, Commentary no. 88, IFS, London, 2002
(www.ifs.org.uk/inequality/childpov.pdf), or M. Brewer, T. Clark and A. Goodman, ‘What
really happened to child poverty in the UK in Labour’s first term?’, Economic Journal,
forthcoming.
48
Child poverty targets and the child tax credit
numerous factors, including growth in earnings and unearned income, as well
as changes in the population, household composition, patterns of employment,
tax and benefit policies, and take-up of means-tested benefits and tax credits.
To make our forecast, we assume that the population, employment rates and
household composition do not change from their 2000–01 values, but that real
earnings grow in line with past trends. (We assume that all workers benefit
from real earnings growth of 2.3% between April 2000 and April 2004, the
average annual rate between 1998 and 2001.) Our model estimates that this
growth in real earnings increases median income – and thus the poverty line –
by an annual average of 1.96%, or 8.05% over four years; real median income
grows by less than real earnings because earnings are not the only source of
income for the median household, and the progressive nature of the income
tax system will tend to make net earnings grow more slowly than gross
earnings.16 Other things being equal, a rise in median income resulting from a
uniform increase in earnings for all workers will tend to worsen child poverty.
This is because earnings are a less important source of income for poor
households with children than they are for the median household, and nonearned income sources (except pre-announced changes in taxes and benefits)
have not been increased in real terms in our forecast.
Table 4.1 shows our overall assessment of likely changes in child poverty
between 2000–01 and 2004–05. Column 1 shows the actual level of child
poverty in 2000–01. Column 2 shows our estimate of the change in child
poverty over the following four years arising solely from the government’s tax
and benefit reforms.17 In other words, adjusting suitably for likely price
changes over that period, it shows how different child poverty would have
been in 2000–01 had the 2004–05 tax and benefit system already been in place
then. We estimate that these reforms to personal tax and benefits could reduce
the number of children in households with incomes below 60% of the median
by 0.8 million.
Column 3 gives our forecast of the impact of three years of real earnings
growth, pushing around 0.2 million children back below the new, higher
poverty line. Column 4 shows the predicted overall impact of both policy
changes and earnings growth. Applying this to the actual child poverty data in
the first column gives us the predicted level in 2004–05 (column 5). The
bottom line is that child poverty is likely to reach 3.3 million children in
2004–05 (AHC), slightly adrift of the government’s target of 3.1 million.
For comparison, we also show estimates for other poverty lines the
government could have chosen relative to median income. If it had stated that
16
Actual average growth in median income in recent years has been higher than 1.96%, but
our estimate is in line with those of other researchers who performed similar calculations (see
D. Piachaud and H. Sutherland, ‘Changing poverty post-1997’, CASEpaper 63, London
School of Economics, Centre for Analysis of Social Exclusion, 2002).
17
This change is actually calculated from an estimate of child poverty in 2000–01 generated
by our model, which simulates tax payments and benefit entitlements rather than drawing on
actual tax payments and benefit receipts as the HBAI methodology does. The model gives
numbers of children in poverty of 2.4 million (50% median), 4.0 million (60% median) and
5.2 million (70% median). Although these starting levels are fractionally different from the
HBAI estimates, the calculated changes can reasonably be applied to the actual figures.
49
Green Budget, January 2003
the target was to be assessed by measuring incomes before housing costs (not
shown in the table), our estimate is that child poverty would reach 2.3 million
children by 2004–05, precisely meeting the target on that definition.18
The estimates in Table 4.1 have attempted to take into account likely real
earnings growth between 2000–01 and 2004–05. But many other things could
affect child poverty over this period, such as the number of families and
children, and the employment rates of parents and other adults. In addition,
our assumptions may be incorrect: for example, earnings growth may differ
between low-paid and high-paid workers (although recent evidence supports
our assumption of uniform earnings growth19), or take-up of means-tested
benefits and tax credits may change. This means that there is a considerable
degree of uncertainty around the estimates in Table 4.1. But, on our central
forecast, the government will need to do more to help low-income families
with children in order to hit its after-housing-costs child poverty target for
2004–05. We quantify how much more in Section 4.4.
Table 4.1. Possible changes in child poverty (AHC), 2000–01 to 2004–05
(millions of children)
Poverty
line
(1)
(2)
(3)
(4)
(5)
Actual child Change due to Change due to Total change, Estimate for
poverty,
tax and benefit
earnings
2000–01 to 2004–05 using
2000–01,
reforms,
growth,
2004–05
HBAI values
HBAI
2000–01 to
2000–01 to
for 2000–01
2004–05
2004–05
2.5
–1.0
+0.1
–0.9
1.7
50%
median
60%
3.9
–0.8
+0.2
–0.6
3.3
median
70%
5.0
–0.6
+0.1
–0.5
4.6
median
Notes: Column 4 = column 2 + column 3; column 5 = column 1 + column 4. Column 3
assumes annual real earnings growth of 2.31%. Numbers are rounded to the nearest 100,000,
as is the custom in HBAI, but this should not be interpreted as a measure of accuracy.
Changes are based on unrounded numbers. Numbers may not sum due to rounding.
Source: Authors’ calculations from IFS tax and benefit model, TAXBEN, using 2000–01
Family Resources Survey.
18
D. Piachaud and H. Sutherland, ‘Changing poverty post-1997’, CASEpaper 63, London
School of Economics, Centre for Analysis of Social Exclusion, 2002, use a similar
methodology. Their results (page 26) are less optimistic than ours, as they estimate that
policies introduced between 2000–01 and 2003–04 might reduce child poverty by 0.6 million
(AHC) – a slightly smaller number than our estimate for policies introduced between 2000–01
and 2004–05 – and that real earnings growth between 2000–01 and 2003–04 might increase
child poverty (AHC) by around 250,000 – a slightly larger number than our estimate of the
impact over the period 2000–01 to 2004–05. The differences in estimates are due to
differences in the data-sets, in the tax and benefit models or in the assumptions about take-up
of benefits and about future real earnings growth.
19
D. Piachaud and H. Sutherland, ‘Changing poverty post-1997’, CASEpaper 63, London
School of Economics, Centre for Analysis of Social Exclusion, 2002, find that earnings
growth at the bottom of the full-time earnings distribution was around as high as that at the
median.
50
Child poverty targets and the child tax credit
Nonetheless, it is important to remember that, if the estimates in Table 4.1 are
realised and the government misses its target, this would still represent a
substantial fall in child poverty from its 1996–97 level. Child poverty would
still be at its lowest level since 1990, having fallen by 1.1 million children
from its peak of 4.4 million in 1996–97.
4.4 What would it cost to reduce child
poverty further?
So how generous does the child tax credit need to be for the government to
achieve its child poverty target in 2004–05? The rates of the child tax credit
for April 2003 have been pre-announced, so the earliest date from which an
increase could take effect is April 2004. We have estimated the number of
children that would be taken out of poverty by various increases in the perchild element of the child tax credit in April 2004 above the default option of
increases in line with earnings growth (which has already been included in the
public finance forecasts). By way of comparison, we have also shown the
implications of increasing child benefit or increasing the family element of the
child tax credit (together with the extra child tax credit for families with babies
in their first year). The results are shown in Table 4.2.
Table 4.2. Effect of possible increases in per-child element of the child tax
credit in April 2004
Increase in per-child element
Weekly per-child
Number of children taken
Cost per
of child tax credit in April
child tax credit
out of poverty (60%
year, 2004
2004
rate, 2004 prices
median income AHC)
prices
(£ p.w.)
(£ p.w.)
(millions)
(£ million)
Average earnings growth +£2
31.05
0.1
660
Average earnings growth +£3
32.05
0.2
1,000
Average earnings growth +£5
34.05
0.4
1,690
Average earnings growth +£10
39.05
1.0
3,460
Other possible changes in
April 2004 (£ p.w.)
£3 on child benefit
n/a
0.2
2,000
£6 on the family element of
n/a
0.2
2,100
child tax credit
Notes: ‘Number of children taken out of poverty’ is rounded to the nearest 100,000 and ‘Cost
per year’ is rounded to the nearest £10 million, but these should not be interpreted as measures
of accuracy. Changes are based on unrounded numbers. The poverty line was allowed to move
if the reform altered median household income.
Source: Authors’ calculations from IFS tax and benefit model, TAXBEN, using 2000–01
Family Resources Survey.
To reduce child poverty from 3.3 million to the target level of 3.1 million, our
estimates suggest that the per-child element of the child tax credit would need
to be increased by around £3 per week in addition to the indexation in line
with earnings growth that has already been promised. This would cost
approximately £1 billion (in 2004 prices). According to our calculations, the
51
Green Budget, January 2003
results from the Piachaud and Sutherland study20 suggest that more
expenditure than this may be required. Equivalent reductions in child poverty
could be achieved by increasing child benefit by £3 per week or by increasing
the family element of the child tax credit by £6 per week. But both these
options would cost around twice as much, because both would benefit well-off
families with children and not just those on low-to-middle incomes.21
Figure 4.4 shows the average percentage income increase in each decile of the
population for three of the options in Table 4.2. We can see that the child
element of the child tax credit is a well-targeted instrument for tackling child
poverty, with the gains concentrated in the bottom half of the income
distribution. (The estimated gains have been averaged over all families,
whether or not they contain children: the bottom decile gains by less than the
second decile because there are fewer children in the bottom decile than in the
second decile.)
Figure 4.4. Distributional effects of different increases in child-related
benefits and tax credits in April 2004
Percentage change in net income
1.5%
1.0%
0.5%
0.0%
Poorest
2
3
4
5
6
7
8
9
Richest
Income decile
£3 on child element of child tax credit
£3 on child benefit
£6 on family element of child tax credit
Notes: The figure shows gains in addition to the increase to the per-child element of the child
tax credit in line with earnings growth in April 2004 to which the government is already
committed. Income deciles are derived by dividing all families (with and without children)
into 10 equally sized groups according to income adjusted for family size using the
McClements equivalence scale. Decile 1 contains the poorest tenth of the population, decile 2
the second poorest and so on, up to decile 10, which contains the richest tenth.
Source: IFS tax and benefit model, TAXBEN, based on 2000–01 Family Resources Survey.
20
D. Piachaud and H. Sutherland, ‘Changing poverty post-1997’, CASEpaper 63, London
School of Economics, Centre for Analysis of Social Exclusion, 2002.
21
All families with children gain from increases in child benefit, and increases in the family
element of the child tax credit by £6 p.w. would benefit all but around half a million families
with children with a combined gross income of over £62,830.
52
Child poverty targets and the child tax credit
It is important, though, to remember that the government may have other
considerations when thinking about increasing further the amount paid per
child. For example, the amount of extra income that parents receive from the
government because they have children has increased substantially since 1997:
the poorest families now receive more for their first child than a young
jobseeking adult is entitled to receive for him or herself.22 Secondly, increases
in the child tax credit also have small but adverse effects on parents’ work
incentives; these could be counteracted with changes to the working tax credit,
but this would increase the cost.23 These factors will need to be considered
alongside the desire to reduce child poverty.
4.5 Conclusion
Despite its commitment to increase the per-child element of the child tax
credit by more than the rate of inflation in April 2004, the government is more
likely than not to miss its target of reducing child poverty by a quarter from its
1998–99 level by 2004–05, on an after-housing-costs basis. Once the likely
effect of future earnings growth on median income and child poverty is taken
into account, around £1 billion of extra spending in 2004–05 may be required
for the government to hit this target, although there are considerable
uncertainties around this estimate. But if this extra money cannot be found,
and if our forecasts for the level of child poverty in 2004–05 are correct, then
the government could still point to a substantial fall in child poverty to its
lowest level since 1990. Furthermore, if a before-housing-costs definition of
child poverty is adopted, our results indicate that the government may indeed
hit its target.
Mike Brewer and Greg Kaplan
22
See S. Adam, M. Brewer and H. Reed, The Benefits of Parenting: Government Financial
Support for Families with Children Since 1975, Commentary no. 91, IFS, London, 2002
(www.ifs.org.uk/taxben/c91.pdf). The weekly jobseeker’s allowance (JSA) for an adult under
25 will be £43.25 from April 2003, compared with £54.25 in child-related payments that will
be received by a non-working family on income support or JSA for their first child – a gap
which will widen as the child element of the child tax credit increases in line with earnings
growth while there are no real increases in JSA.
23
There are two impacts: increases in any part of the child tax credit reduce the need for
parents to work; and increases in the per-child element of the child tax credit widen the range
of incomes over which parents can face a marginal deduction rate of 70% (they only receive
30p extra income for each £1 of extra earnings), which reduces the cash gain from working.
53
5. Income tax and National
Insurance contributions
The Labour manifesto for the 2001 general election promised that ‘we will not
raise the basic or top rates of income tax in the next Parliament’.1 This pledge
featured heavily in the party’s successful re-election campaign (and, indeed,
the previous election campaign of 1997). However, the government refused to
rule out increases in other taxes if re-elected, and in his first Budget Speech of
the new parliament in April 2002, Chancellor Gordon Brown announced that
National Insurance contributions (NICs) for employees, employers and the
self-employed would go up with effect from April 2003.
The government estimates that the National Insurance (NI) increases will raise
around £8.2 billion in 2003–04.2 The Chancellor stated that the increases were
necessary to fund the increases in NHS spending also announced in the April
2002 Budget. But increases in NICs of this magnitude, when set against the
commitment not to raise income tax, raise several questions. Why did the
government feel it necessary to make a pledge on income tax but not on NI?
What is the justification for having NI as a separate tax from income tax?
Have the two taxes become more alike? And how might the two systems
develop in future?
In this chapter, we start by briefly discussing the historical relationship
between NI and income tax. We then compare and contrast the current income
tax system with the current NI regime and look at the distributional effects of
the changes scheduled for April 2003. We finish by assessing what further
reforms of the NI and income tax systems might be likely, and what might be
economically desirable.
5.1 Income tax and National Insurance: a
history of convergence
Income tax and National Insurance contributions have traditionally played
quite separate roles in the tax system. Income tax was introduced in 1799 to
raise revenue for general government expenditure. The present structure of the
income tax system came into being in 1973: taxpayers have a tax-free personal
allowance, and income above this level is taxed at progressively higher rates.
In contrast, National Insurance started life in 1948 as a social insurance
scheme based on the ‘contributory principle’ – that benefits received should
1
Page 10 of Labour Party, Ambitions for Britain (Labour’s manifesto 2001), London, 2001
(www.labour.org.uk/ENG1.pdf).
2
Source: authors’ calculations from HM Treasury, Financial Statement and Budget Report,
HC592, Stationery Office, London, 2002 (www.hm-treasury.gov.uk/budget/bud_index.cfm)
and HM Treasury, Tax Ready Reckoner and Tax Reliefs, London, 2002 (www.hmtreasury.gov.uk/mediastore/otherfiles/adtrr02.pdf).
54
Income tax and National Insurance contributions
reflect contributions paid. Workers and their employers paid contributions at a
flat rate, independent of earnings, in return for entitlement to various flat-rate
benefits. 1961 saw a major change as NICs became earnings-related for the
first time. Since benefits received bore little relation to previous earnings, this
constituted a weakening of the contributory principle and made NICs more
like income tax. Since then, the two systems have moved closer together in
many more ways – largely through reforms that have made NI operate more
like income tax. Under the current system, NICs are levied as a percentage of
employees’ earnings and employers’ labour costs above a lower weekly
earnings limit. Box 5.1 gives an explanation of the jargon associated with the
two systems.
Box 5.1. A glossary of income tax and National Insurance terms
The personal allowance is the income on which no income tax is paid. In
2003–04, it will be £4,615 per year, or £89 per week (higher for those aged 65
or over). Income immediately above this is taxed at the starting rate, currently
10%.
The basic-rate threshold is the level of taxable income (i.e. income above the
personal allowance) at which the 10% rate stops and the basic rate of tax,
currently 22%, starts to be paid. The threshold for 2003–04 will be announced
in the next Budget, but the ‘default’ increase (in line with inflation) would set
it at £1,960 per year, so that those with incomes above £6,575 per year, or
£126 per week, would pay basic-rate tax.
Similarly, the higher-rate threshold is the level of taxable income at which the
40% higher rate of tax becomes payable. Default indexation would take the
higher-rate threshold to around £30,500 in 2003–04; higher-rate taxpayers
would be those with incomes above £35,115 per year, or £675 per week.
The lower earnings limit (LEL) is the level of earnings – £77 per week in
2003–04 – at which employees build up entitlement to NI (contributory)
benefits. Employees do not actually start paying contributions, however, until
the primary threshold (PT) is reached; employer contributions begin at the
secondary threshold (ST) and self-employed contributions at the lower profits
limit (LPL). At the moment, the PT, ST and LPL are all equal to the income
tax personal allowance, i.e. £89 per week in 2003–04.
Until April 2003, the upper earnings limit (UEL) and upper profits limit
(UPL), both £595 per week in 2003–04, are the level of earnings at which
employees and the self-employed respectively stop paying contributions
(employer contributions have no limit). From April 2003, however,
contributions of 1% will be payable above the UEL and UPL.
The following are some of the most important reforms that have contributed to
the convergence of NI and income tax:
•
In 1990, the income tax system moved from a joint system of assessment
(where a married woman’s income was treated as her husband’s income for
tax purposes) to an individual system, where both partners pay tax
separately. This moved income tax closer to the NI system, which is based
55
Green Budget, January 2003
on individual earnings. Further to this, the married couple’s income tax
allowance was abolished in 2000.3
•
The levels of weekly earnings at which employees and employers start
paying NICs were aligned with the weekly level of the income tax personal
allowance4 in 1999 (for employer contributions) and 2001 (for employee
contributions).
•
Both employee and employer NICs used to have a ‘kink’ in the
contributions schedule whereby, as weekly earnings passed the lower
earnings threshold, contributions became payable on the whole of weekly
earnings, not just earnings above the threshold. This kink was finally
eliminated in 1999; contributions are now payable only on earnings above
the threshold. This is equivalent to the treatment of income in the income
tax system.
•
The tax rates levied on the majority of taxpayers under the two systems
have moved closer together. The basic rate of income tax has decreased
from 33% in 1979 to 22% in 2002–03. Meanwhile, the rate of employee
NICs rose from 6.5% to 10% between 1979 and 2002; it will rise to 11% in
2003. The standard rate of employer NICs rose from 10% to 11.8%
between 1979 and 2002; it will rise to 12.8% in 2003.5
•
The treatments of those with higher incomes have also become more
similar. Before 1985, NICs were not payable on earnings above the weekly
upper earnings limit. This was in contrast to income tax, where there was
(and is) no such upper limit. However, the UEL on employer NICs was
removed in 1985. For employee NICs, the UEL is still in place, but the
additional 1 percentage point on employee NICs from April 2003 will
apply to earnings above the UEL as well as earnings below it. Meanwhile,
the series of progressively higher income tax rates (ranging from 40% to
83%) that existed before 1988 have given way to a single 40% higher rate.
•
The contributory principle of NI has been eroded over time. Not only are
contributions increasingly earnings-related, as discussed above, but also
benefits are increasingly related to current circumstances rather than past
contributions. Most of the main benefits and tax credits are noncontributory, relying instead on means testing (e.g. housing benefit,
working families’ tax credit, income support and the minimum income
guarantee), and their generosity has increased in recent years. Contributory
benefits such as the basic state pension and contributory jobseeker’s
allowance, on the other hand, have mostly been frozen in real terms since
the 1980s. Expenditure on contributory benefits has therefore fallen as a
proportion of total government benefit (and tax credit) expenditure, from a
3
The introduction of the children’s tax credit in 2001 reintroduced some joint assessment into
the income tax system. The married couple’s allowance still exists for people born before 6
April 1935.
4
By ‘weekly level’, we mean the weekly equivalent, for someone working throughout the
year, of the level of the annual income tax personal allowance.
5
The employer and employee contribution rates given here apply to individuals contracted
into the State Second Pension (S2P).
56
Income tax and National Insurance contributions
high of 76% in 1965–66 to an estimated 45% in 2003–04.6 Furthermore,
some so-called ‘contributory’ benefits can now be received by people who
have not actually paid contributions: for example, since 1999, people with
earnings between the LEL and the PT receive benefit entitlements despite
not paying contributions; and since 2001, incapacity benefit has been
available to non-contributors if they have been unable to work from a
young age. Conversely, some ‘contributory’ benefits are not available to
people who have contributed (since 2001, for example, incapacity benefit
has been means-tested against an individual’s private pension income).
Taken as a whole, the changes to NICs and the income tax system detailed
above have made the two systems much more similar. Nonetheless, there
remain substantial differences between the two systems. In the next section,
we look at these differences and examine what the combined structure means
for effective marginal tax rates. We also take a detailed look at the
distributional effect of the recently announced NI increases.
5.2 The system from April 2003
Despite the convergence of income tax and NICs over the years, there remain
differences between them. The most obvious difference is in the rates and
thresholds, shown in Table 5.1 (Box 5.2 lists the reforms announced in the
2002 Budget that take effect in April 2003). Both income tax and NICs
become payable at the same level – £89 per week, or £4,615 per year – but
while income tax has increasing marginal rates, the rate of NICs falls at higher
levels of earnings.
Table 5.1. Rates of income tax and National Insurance contributions,
2003–04
Income taxa
Weekly
Tax rate
incomeb
(%)
National Insurance contributions
Employee
Employer
Selfrate (%)
rate (%)
employed
rate (%)c
£0–£89
0
0
0
£89–£595
11f
12.8f
8
£595–
1
12.8
1
Weekly
earnings
£0–£89d
0
£89–£126e
10
£126–£675e
22
£675–e
40
a
Gross of tax credits.
b
Income tax is an annually-based system; this table therefore assumes year-round work.
c
The self-employed also pay flat-rate contributions of £2 per week if their profits exceed £79.
d
A higher tax-free allowance applies for those aged 65 or over with all thresholds moved up
correspondingly; the extra allowance is tapered away at higher income levels.
e
Assumes 1.7% indexation of basic- and higher-rate income tax thresholds and statutory
rounding.
f
If an employee is contracted out of the State Second Pension, a reduced rate applies.
6
Source: authors’ calculations from Department for Work and Pensions, ‘Benefit Expenditure
Tables’ (www.dwp.gov.uk/asd/asd4/expenditure.htm).
57
Green Budget, January 2003
Box 5.2. Reforms to income tax and National Insurance contributions
taking effect in April 2003
•
The employer rate of NICs will rise by 1 percentage point.
•
The employee rate of NICs will rise by 1 percentage point. This
increase will extend to earnings above the UEL, which were not
previously subject to employee NICs.
•
The self-employed rate of NICs will rise by 1 percentage point. This
increase will extend to profits above the UPL, which were not
previously subject to NICs.
•
The PT, ST and UPL – the level at which NICs start to be paid – will
be frozen at £89 per week, and so fall in real terms.
•
The income tax personal allowance will be frozen at £4,615 in 2003–
04, and so fall in real terms.
Figure 5.1. Combined payroll tax schedule, 2003–04
60%
Marginal tax rate
50%
40%
30%
20%
10%
0%
0
200
400
600
800
1000
Pre-tax earnings (£/week)
Notes: Assumes 1.7% indexation of basic- and higher-rate income tax thresholds and statutory
rounding. Rates shown are for a childless employee under 60 years old, not contracted out of
the State Second Pension, with no unearned income.
In so far as the main difference between income tax and NICs is the rate
structure, we can simply add up the rates at different earnings levels to find a
combined ‘payroll tax’ schedule; this is done in Figure 5.1, which combines
income tax, employer NICs and employee NICs.7
7
For Figure 5.1 and what follows, we assume that employees bear the full burden of both
employee and employer NICs. The justification for this assumption is discussed in detail in
Chapter 9.
58
Income tax and National Insurance contributions
Yet this is not quite the whole story: other differences remain. One is that
income tax is assessed on an annual basis whereas NI is a weekly system. This
may be important for people working less than a full year.
Another difference is the treatment of the self-employed. Earnings from selfemployment are treated like any other earnings for income tax purposes. NICs,
however, are much lower for the self-employed, as Table 5.1 shows: the
contribution rate is lower than for employees, and there is no equivalent to the
employer element. The self-employed do pay an additional flat-rate
contribution of £2 per week and also have reduced benefit entitlement. But
even after accounting for that, the government calculates that, in 2001–02, NI
for the self-employed raised £2.4 billion less than it would have done had the
Class 1 (employer/employee) system applied.8
The final difference between income tax and NICs is the tax base: NICs are a
tax on earnings, income tax a tax on income. Income tax, then, has a broader
base: it is payable on unearned income. The biggest single source of unearned
income is pension income, but it also includes property income, interest from
bank accounts, share dividends and many state benefits. These are subject to
income tax but not to NICs.9
Figure 5.1, then, is not a complete representation of income taxation for
anyone with unearned or self-employment income: the combined tax rates for
these will be lower, since less (or no) NICs are paid. Despite these differences,
income tax and NICs are now quite similar. By way of illustration, Figure 5.2
shows the distributional impact of the changes to NICs announced in the 2002
Budget – a freeze in the threshold at which NICs start to be paid, and an extra
1 percentage point on contributions for employers, employees and the selfemployed (including on earnings above the UEL and UPL). Also shown is the
impact of an equivalent change to income tax – a freeze in the personal
allowance and an extra 2 percentage points on the lower, basic and higher
rates.
As Figure 5.2 shows, the impacts of these two sets of reforms are similar,
although not identical. The income tax rise would cost families more
throughout the income distribution – we estimate that it would raise around
£1.7 billion more than the reforms to NICs. The main reason is the taxation of
unearned income – most of the extra revenue would come from people with
higher incomes (and the richest tenth in particular), who not only have more
unearned income but also pay tax on it at a higher rate.10
8
Source: HM Treasury, Financial Statement and Budget Report, HC592, Stationery Office,
London, 2002 (www.hm-treasury.gov.uk/budget/bud_index.cfm).
9
The two systems also differ in their treatment of pension contributions: NICs are not paid on
employers’ pension contributions but are on individuals’; income tax, by contrast, is not paid
on any contributions.
10
Unearned income of higher-rate taxpayers is taxed at the higher rate. Basic-rate taxpayers,
however, pay tax on their income from savings at a slightly reduced rate of 20%; we also
increase this rate by 2 percentage points for this simulation, though it does not make a large
difference to the overall results.
59
Green Budget, January 2003
Figure 5.2. Losses across the income distribution from National Insurance
changes announced in the 2002 Budget, and from a package of similar
changes to income tax
Percentage change in net income
0.0
-0.5
-1.0
-1.5
-2.0
-2.5
-3.0
Poorest
2
3
4
5
6
7
8
9
Richest
Income decile
NI changes
Income tax changes
Notes: Income deciles are derived by dividing all families into 10 equally sized groups
according to income adjusted for family size using the McClements equivalence scale. Decile
1 contains the poorest tenth of the population, decile 2 the second poorest and so on, up to
decile 10, which contains the richest tenth.
Source: IFS tax and benefit model, TAXBEN, run using data from the Family Resources
Survey 2000–01.
The income tax reforms would also raise more money because of the different
treatment of the self-employed. The self-employed will see a 1 percentage
point rise in NICs, whereas they would have faced a 2 percentage point rise in
their income tax rates had that option been chosen. Employees, by contrast,
are affected by both the rise in the employee rate and the rise in the employer
rate, and so pay 2 percentage points more in either case. The self-employed
are disproportionately represented in the top income decile, which also helps
explain why the income tax increase falls more heavily on the top decile than
does the NI increase.
5.3 Options for further reform
As discussed in Chapter 3, the government may feel it has to raise taxes
further in due course to strengthen the public finances. If it chooses to do this
by raising payroll taxes, there are a number of routes it could take.11
One obvious option is a straightforward increase in NI rates of the kind
announced in the 2002 Budget. The distributional impact of such a rise would
11
It should be noted that changes to NI could not take effect immediately – a result of its
weekly structure. Changes could, however, be pre-announced, as they were in the 2002
Budget.
60
Income tax and National Insurance contributions
be similar to that shown in Figure 5.2. An uncapped 1 percentage point rise in
employee NI would raise about £3.8 billion.12 The innovation in the last
Budget of levying 1% employee NICs on earnings above the UEL also creates
a precedent: there is ample scope for increases in this new ‘additional rate’ of
NICs. A 1 percentage point increase above the UEL (and UPL) alone would
raise around £0.8 billion, with the burden of the increase falling mainly on
families in the top income decile. Abolishing the UEL (and UPL) altogether
would be roughly equivalent to a 10 percentage point increase in the
‘additional’ rate, and would raise around £8.3 billion.13
Rises in income tax rates seem much less likely, primarily because of
Labour’s manifesto pledge not to raise the basic or higher rates. The
government could still raise money through income tax, however, by freezing
or even reducing tax allowances and thresholds.
The 2002 Budget froze the personal allowance, and also the PT and ST to
keep them all aligned in 2003–04. The Chancellor might plausibly do this
again in 2004–05, raising about £0.8 billion. Most of the revenue would come
from higher earners (since income tax allowances reduce the amount of
income taxed at the highest rate paid by each taxpayer), but middle-income
families would lose most as a proportion of income.
Reducing the basic-rate threshold seems unlikely because of a manifesto
pledge that ‘we will extend the 10p tax band’.14 That leaves the higher-rate
threshold and the National Insurance UEL. At present, the higher-rate
threshold is higher than the UEL. This causes a dip in the effective tax rate for
income between £595 and £675 per week (as shown in Figure 5.1) for which it
is hard to find an economic rationale. Aligning the UEL and the higher-rate
threshold would correct this anomaly, and would be consistent with previous
changes (such as the alignment of the NI thresholds with the income tax
personal allowance). Indeed, the Chancellor moved in this direction by
increasing the UEL by more than inflation in both April 2000 and April 2001,
thus reducing the gap to the higher-rate threshold.
An alignment could be achieved either by raising the UEL to match the
higher-rate threshold (so that income currently between the two would be
subject to ‘standard’ NICs and basic-rate income tax, like the income below it)
or by lowering the higher-rate threshold to match the UEL (so that income
currently between the two would be subject to ‘additional’ NICs and higherrate income tax, like the income above it). Either reform would be extremely
progressive – the richest 10 per cent would provide two-thirds of the revenue,
while the bottom half of the income distribution would be virtually unaffected
– but they would raise very different amounts of revenue: increasing the UEL
12
A 1 percentage point rise in employer NI would raise slightly more – about £4 billion –
because, unlike employee NI, it is payable in respect of employees who are at or above the
state pension age.
13
We assume that the UEL would be maintained (at its current level) in its role as a cap on the
band of income where the contracted-out rebate applies, as it has been for employer NICs.
14
Page 10 of Labour Party, Ambitions for Britain (Labour’s manifesto 2001), London, 2001
(www.labour.org.uk/ENG1.pdf).
61
Green Budget, January 2003
to the higher-rate threshold would raise around £1 billion, while reducing the
higher-rate threshold to the UEL would raise a little over £2 billion.
The increasing similarity of income tax and NICs outlined in this chapter
makes this discussion of future changes look rather strange from an economic
perspective. Why, if the two are so similar, are increases in NICs widely
perceived as more likely than increases in income tax? One answer is that
Labour has a manifesto commitment not to raise income tax but no such
commitment in respect of NICs. But that is, at best, a partial answer: why did
Labour feel impelled to make a pledge in respect of one but not the other?
Clearly, the government either perceives the two taxes differently or thinks
that the public perceives them differently. From a purely economic
perspective, it makes little sense to be implacably opposed to a percentage
point increase in the basic and higher rates of income tax, but at the same time
to have no objection to a National Insurance increase with similar effects over
most of the income distribution. The trend towards increased levels of NICs
coupled with lower rates of income tax began almost thirty years ago and
shows no sign of abating. What this demonstrates is that a pledge by the
government or the opposition not to increase basic- or higher-rate income tax
means very little in economic terms, as without a pledge not to increase NICs
either, there is no barrier to the overall level of taxation on earnings increasing
in the future.
On economic grounds, it would seem sensible in many ways to aim towards a
complete integration of the income tax and NI systems. This would offer the
advantages of transparency and administrative efficiency with few apparent
drawbacks. The original rationale for separate systems – the ‘contributory
principle’ underlying the NI system – is all but obsolete. There is no reason to
suppose that the slow death of the contributory principle will reverse, or even
halt, in the near future. For example, the basic state pension (easily the biggest
contributory benefit) is set to increase in line with prices, whereas the
government’s stated long-term aspiration is to increase the non-contributory
minimum income guarantee (MIG) for those aged 60 or over in line with
earnings. That means the basic state pension is due to become less important
relative to the MIG, and contributions records will gradually become less
relevant in determining the amount of benefits received by those aged 60 or
over.
Nor do the other remaining differences between income tax and NI provide a
strong justification for maintaining separate systems. Lower rates of taxation
on unearned income and on self-employment income create a distortion in
favour of these income sources, which is probably undesirable. Yet even if the
government wishes to retain either or both of these anomalies, it would still be
possible to do so within the income tax system (to have a lower rate of tax on
pension income, for example). If the government believes that payroll taxes
have different effects depending on whether they are levied on the employer or
the employee, another option would be to integrate employee NI with income
tax, leaving NI as a pure employer tax.
Given these arguments, the barrier to full integration of the two systems seems
primarily political. Governments appear to believe that raising NICs is likely
to be less costly to them in votes than raising income tax. It will be interesting
62
Income tax and National Insurance contributions
to see if the high-profile increases in NICs due to take effect in April change
this political calculus. If they do not, full integration of NI and income tax
remains much less likely than further gradual alignment, with equalisation of
the UEL and higher-rate threshold perhaps the obvious next step.
Stuart Adam and Howard Reed
63
6. Company taxation and innovation
policy
This chapter begins with a discussion of four issues in company taxation. First,
we assess the recent government consultation document on reform to
corporation tax, which looks at possible changes to the rules for calculating
taxable income. (This follows a series of reforms to the taxation of corporate
profits since 1997, which are discussed in Chapter 9.) In Section 6.2, we look
at the taxation of dividends, in the light of the changes to the UK tax system
since 1997 and recent proposals for reform in the USA. In Section 6.3, we
assess the structure of North Sea taxation following the changes announced in
Budget 2002 and in the 2002 Pre-Budget Report. Then, in Section 6.4, we
consider an issue that is not on the government’s immediate agenda, but one
where there might be a case for reform – stamp duty on share transactions.
The chapter also looks at innovation policy, following the 2002 Pre-Budget
Report announcement of a review into the interaction between universities and
business, which will report at the same time as a separate review of the UK’s
innovation performance. In the final section of the chapter, we examine trends
in UK research and development (R&D) activity and consider current policy
towards innovation, including the two new R&D tax credits.
6.1 The August 2002 consultation
In August 2002, the government issued a consultation document on further
reform to the corporation tax system.1 This considered possible reforms in
three main areas relating to the calculation of taxable income: the treatment of
depreciation and gains and losses on capital assets; the schedular system,
which distinguishes between income from different sources; and the
distinction between trading companies and investment companies.
The objective of the proposed reforms in each area is to align the calculation
of taxable income more closely with the measurement of profits in company
accounts. This follows the approach taken in recent changes to the taxation of
intangible assets (Finance Act 2000), loan relationships (Finance Act 1996)
and foreign exchange gains and losses (Finance Act 1993). In many areas,
such alignment with accounting practice can achieve useful simplification of
the tax system. However, some of the proposals considered in the consultation
document could have radical impacts on tax reliefs for depreciation and on the
tax treatment of losses, as we discuss in this section.2 As the trading/
1
HM Treasury and Inland Revenue, Reform of Corporation Tax, London, 2002
(www.inlandrevenue.gov.uk/consult_new/taxreform_final.pdf).
2
A more detailed response to this consultation is provided by A. Klemm and J. McCrae,
Reform of Corporation Tax: A Response to the Government’s Consultation Document,
Briefing Note no. 30, IFS, London, 2002 (www.ifs.org.uk/corptax/bn30.pdf). A wider
discussion of the issues involved in aligning tax and commercial accounts is provided by G.
64
Company taxation and innovation policy
investment company distinction affects only a limited number of firms, our
discussion here focuses on the first two issues.
Capital allowances
Capital equipment that is used in producing goods or supplying services
typically falls in value as it is used, and may become worthless at the end of its
useful life. This fall in the value of capital assets, known as economic
depreciation, represents a cost to the owner, which is reflected in depreciation
charges against profits in company accounts, and in capital allowances against
taxable income in the corporation tax. Capital allowances claimed in 1999–
2000 amounted to £65 billion, or about half of total income chargeable to
corporation tax (net of these and all other allowances).3 About 90% of these
capital allowances were for plant and machinery. Any reform to the system of
capital allowances could therefore have large effects on the tax bills facing
individual companies, even if it were revenue-neutral overall. Capitalintensive industries such as energy, engineering and other manufacturing
could be particularly affected.
Under the current rules, most types of plant and machinery qualify for a 25%
per annum writing-down allowance, on a declining-balance basis.4 There are
more generous provisions for favoured forms of investment, such as some
energy-saving technologies and investment by smaller companies, and a lower
rate for some ‘long-lived’ assets used mainly by utility companies. Industrial
buildings benefit from a 4% per annum writing-down allowance on a straightline basis,5 whilst there is no writing-down allowance for commercial
buildings.
At present, there is no connection between these writing-down allowances
prescribed for tax purposes and the depreciation charges that are used in firms’
published accounts. In reporting their profits, firms may use higher or lower
depreciation rates, depending on the nature of the capital assets they own. For
example, expenditure on computers with very short useful lives may be
written off more quickly, whilst expenditure on pipelines with very long useful
lives may be written off more slowly. The treatment of these assets in
company accounts has no implications for the firm’s corporation tax liability.
Replacing the current system of capital allowances by a deduction for the
depreciation charge reported in company accounts would have major
implications for firms that account for depreciation at substantially different
rates from the current schedules of capital allowances. Retailers who
depreciate their commercial property would benefit, gaining a deduction for
depreciation where none is currently provided under capital allowances.
Macdonald, The Taxation of Business Income: Aligning Taxable Income with Accounting
Income, Tax Law Review Committee Discussion Paper no. 2, IFS, London, 2002.
3
Inland Revenue Statistics (www.inlandrevenue.gov.uk/stats/index.htm).
4
For a £100 investment, the capital allowance is £25 in the first year, £18.75 in the second
year (i.e. 25% of the remaining £75 balance), and so on. These allowances are treated as costs
in the calculation of taxable income for each period.
5
For a £100 investment, the capital allowance is £4 per year for a period of 25 years.
65
Green Budget, January 2003
Engineering firms that write off their plant and machinery more slowly than
the 25% capital allowance rate would tend to lose out, receiving less generous
depreciation deductions than under the current tax rules.
The consultation document omits any detailed analysis of which sectors would
gain and lose from this change, whether the tax relief for depreciation would
become more or less generous overall as a result, and what the consequent
effects on corporation tax revenues would be. We estimate that, on average,
depreciation rates used in company accounts tend to be lower than those
specified by capital allowances. If so, tax relief for depreciation would become
less generous, on average, if this reform were to be implemented, and
corporation tax payments would tend to be higher, unless there were an
offsetting reduction in the corporation tax rate. There would also be substantial
redistribution of corporation tax payments, with plant-and-machineryintensive sectors such as manufacturing tending to lose and other sectors, such
as services, tending to gain.6
Such a change in the tax treatment of depreciation would have a substantial
impact on the cost of capital and the incentive to undertake investment in the
UK. Existing capital allowances do not provide particularly generous tax relief
for depreciation when compared with corporate income taxes in other major
economies.7 Whilst there may be a case for linking tax depreciation schedules
more closely to economic depreciation rates for assets used by different
sectors, or for cutting tax allowances for depreciation overall, such radical
reforms should clearly be evaluated with regard to their effects on business
investment and corporation tax payments, and not slipped in under the banner
of simplifying the calculation of taxable income.
A problem with aligning tax and accounting depreciation deductions is that, by
choosing to increase the rate at which capital expenditures are written off
against profits, firms would be able to defer their corporation tax payments.
Not only would this be a potential threat to corporation tax revenues in the
longer term, but it would also reduce the quality of the information provided in
published accounts. Especially after recent accounting scandals, it would seem
risky to introduce tax incentives for companies to produce accounts that may
not provide an appropriate measure of profits. To avoid too much abuse, the
likely outcome would be prescribed maximum rates of depreciation for
different types of assets, similar to the present capital allowances, with
accounting depreciation provisions tending to converge on these prescribed
rates. So rather than aligning tax allowances with company accounts, the result
may be the opposite, i.e. to align accounts with tax rules.
6
A. Klemm and J. McCrae, Reform of Corporation Tax: A Response to the Government’s
Consultation Document, Briefing Note no. 30, IFS, London, 2002
(www.ifs.org.uk/corptax/bn30.pdf), discuss the available empirical evidence in more detail.
7
A. Klemm and J. McCrae, Reform of Corporation Tax: A Response to the Government’s
Consultation Document, Briefing Note no. 30, IFS, London, 2002
(www.ifs.org.uk/corptax/bn30.pdf), compare the tax treatments of depreciation in 16 OECD
countries.
66
Company taxation and innovation policy
The schedular system
The consultation document also considers changes to the schedular system
currently used to compute taxable income, which would have implications for
the tax treatment of losses. Like most corporate income taxes, corporation tax
is charged when taxable income is positive, but there is not a symmetric
treatment when taxable income is negative. A symmetric treatment would
require a negative tax payment, or a payment from the Inland Revenue to the
firm, equal to the tax rate times the loss. Instead, the loss can, at best, be set
against taxable profits from a limited number of earlier years. When this carryback provision is exhausted, losses can only be carried forward to set against
taxable profits in subsequent years, with no compensation for the delay before
losses can actually be used.
These limitations on loss relief discriminate against large, risky investments,
which, in the event that they turn out to be unsuccessful, could push the firm
into a loss-making position. High tax payments expected if returns turn out to
be high are not balanced by tax rebates expected if profits turn out to be
negative, an effect that can be avoided if the firm chooses a safer investment
with less chance of returning a loss. This effect is particularly important in the
case of new, start-up firms, which do not have past profits against which
losses can be carried back, and which may not expect to generate positive
taxable profits for some considerable time. The tax treatment of the same
investment project undertaken by an established firm with profits generated by
existing operations is also more generous, which runs counter to the
government’s objective of stimulating enterprise and business start-ups.
The schedular system in the UK corporation tax introduces further limitations
on relief for losses. Profits and losses from different sources are not
aggregated at the level of the firm, but are calculated separately under
different ‘schedules’, and ‘cases’ within schedules (for example, Schedule A
Property Income and Schedule D Case I Trading Income), and for trading
income, separately for different trades. In the current year, any losses can
generally be offset against current profits from any schedule or trade. Losses
carried back or forward, however, can generally only be offset against profits
from the same schedule or trade.8 The result is that a loss generated on one
activity may not even be offset against a profit generated by the same firm but
from a different source and in a different time period. Such losses in an
otherwise profitable firm are often called trapped losses, particularly if they
stem from a scaled-down or abandoned activity that is not expected ever to
produce profits high enough to relieve past losses. The effect of the schedular
system is that integrated companies are currently taxed similarly to groups, as
group relief also only allows losses to be offset against profits of other
subsidiaries in the same accounting period. The origins of the schedules,
however, date back to the development of the personal income tax in the
nineteenth century, and have no parallel in the calculation of profits in
company accounts.
8
The exact limitations on the set-off of losses vary somewhat across different schedules and
cases.
67
Green Budget, January 2003
The Inland Revenue estimates that, of the £80 billion losses generated by UK
companies in 2000–01, around £15 billion were unrelieved and left to be
carried forward to later years. Providing full relief for these losses would
therefore be expensive: using the standard corporation tax rate of 30%
suggests a cost of around £4.5 billion, compared with total corporation tax
receipts in 2000–01 of around £32 billion. Such expensive measures are
unlikely to be on the government’s agenda at the present time.
The consultation document does not estimate what fraction of these losses
would be relieved under various proposed relaxations of the schedular system,
but it is likely to be rather small. Of much greater quantitative significance
would be changes to the current system of group relief. If the schedular system
were to be significantly relaxed or abolished, then there would be a strong case
for relaxing the limits on group relief as well, allowing group-wide carryforward of losses. Otherwise, there would be a tax incentive for groups with
large stocks of losses carried forward to integrate their subsidiaries, so as to be
able to offset these losses against profits from other activities. Another
argument for relaxing the limits on group relief would be the logic of aligning
taxable income more closely with accounting profit, which would suggest
taxing groups of companies on the same consolidated basis as reported in their
consolidated accounts. However, the cost of the additional loss relief that this
implies may make such changes prohibitively expensive for the foreseeable
future.
The extent to which changes to the schedular system would deal with the
concern over start-up companies is unclear, since relatively few new firms are
likely to have income from more than one source. A more targeted approach to
this issue would be to allow at least some tax reliefs to be paid immediately to
loss-making firms in their start-up phase, along the lines of the R&D tax credit
for small and medium-sized companies introduced in April 2000.
6.2 Dividend taxes
Since 1997, there have been important changes to the taxation of dividends in
the UK – namely, the abolition of repayable tax credits to tax-exempt
shareholders in 1997 and the abolition of advance corporation tax (ACT) in
1999. While no new reforms have been announced since the abolition of ACT,
this topic remains interesting, especially as the USA has just announced a
possible change to its system of dividend taxation.
Background
Across the world, many different approaches to dividend taxation are taken.
This is because different views can be held about the extent to which double
taxation of dividends should be avoided. Double taxation can arise because
dividends are paid out of taxed profits but may then be subject to income taxes
levied on the recipient of the dividend.
In a classical system of dividend taxation, corporate taxes and taxes paid by
recipients of dividends are completely separate, and dividends therefore face
68
Company taxation and innovation policy
taxation both at the firm and at the individual level. Theoretically, there are
two ways this double taxation can be completely avoided. One possibility is to
tax dividends only at the shareholder level. This can be achieved by a full
imputation tax system, in which a tax credit accompanies dividends paid out
of taxed profits, reflecting the full amount of corporate tax paid on the
underlying profits. The individual can then set this tax credit against his or her
own tax liability. If the personal tax rate is higher than the corporation tax rate,
additional tax payments are required; if it is lower, some of the tax credit will
be paid out to the shareholder. The other possibility is to tax dividends at the
company level only. This can be achieved by exempting dividends from
income taxation, which is in fact the US proposal. Whilst full imputation
ensures that dividends are taxed only once at the recipient’s marginal income
tax rate, the latter approach imposes a single flat-rate charge at the corporate
income tax rate.
In practice, few countries employ any of these systems in a pure form. Instead,
most countries operate systems that relieve some, but not all, of the possible
double taxation of dividends. Systems practised include: applying a lower
personal tax rate on dividends than on other personal income; using partial
imputation (i.e. dividends come with a tax credit, but the tax credit does not
reflect the full tax paid at the corporate level); or applying a lower corporation
tax rate on distributed earnings. The systems of some countries incorporate
more than one of these features.
The system in the UK
From 1973 to 1997, the UK used a partial imputation system, in which
dividends were accompanied by a tax credit that was set at the standard
personal income tax rate, so that basic-rate taxpayers did not face any
additional tax liability on dividend income. Higher-rate taxpayers had to pay
additional tax on dividends, and tax-exempt shareholders could claim back the
tax credit. The latter point was especially important for pension funds, as a
high proportion of shares are held indirectly through such institutions. The
reason for calling this system a partial imputation system is that the rate of the
tax credit was generally lower than the corporation tax rate.
Since the July 1997 Budget, these tax credits are no longer repayable to taxexempt institutional shareholders. For other domestic shareholders, nothing of
substance changed,9 so that the UK now has a hybrid system, with partial
imputation for taxpaying shareholders and a classical treatment of tax-exempt
shareholders. The reform could therefore be seen as a move towards a more
classical system of dividend taxation. In effect, the current UK system could
also be described as having no income tax on dividends for most domestic
shareholders and a preferential tax rate on dividends for higher-rate taxpayers.
9
The tax credit was cut from 20% to 10%, but this did not affect domestic shareholders,
because their tax rates on dividend income were reduced correspondingly (currently 10% for
basic-rate taxpayers and 32.5% for higher-rate taxpayers).
69
Green Budget, January 2003
The system in the USA
The situation in the USA is rather different. As yet, the USA is one of only
four OECD countries applying a pure classical system of dividend taxation.10
The recent proposal to abolish dividend taxation at the shareholder level
would therefore move the US system more in line with those of other
countries, which give some relief for double taxation, even if in very diverse
ways. The direction of change in the proposed US reform, away from a
classical system, is opposite to the direction of the recent changes in the UK.
The final system achieved is, however, rather similar. The only substantial
difference in its effect is that the tax liability in the UK is still to some extent
determined by the tax rate of the shareholder, as higher-rate taxpayers face an
additional tax charge on dividends. The US proposals would abandon
progressive taxation of dividends, as dividend income would be subject to a
flat-rate charge at the corporate tax rate. The US announcement also contained
a proposal to exempt retained profits from capital gains taxation at the
personal level. As the practicality of this proposal is somewhat unclear, we do
not discuss it further in this section.
Effects of reforms
The reforms in the UK and those proposed in the USA have both been
accompanied by claims that they would have beneficial effects on investment.
In the USA, the further claim was made that they would be beneficial for the
stock market. Interestingly, even though the reforms go in opposite directions,
it has been claimed in both cases that they will lead to higher investment.
The argument in the UK was that pension funds were thought to prefer a large
share of profits to be paid out rather than reinvested, because only dividends
paid out benefited from the tax credit. It was then argued that removing the
repayability of the tax credit would diminish the pressure that pension funds
were thought to exert on companies to pay out dividends rather than reinvest
earnings, possibly to the detriment of investment. If this were right, the US
proposals would seem unlikely to lead to higher investment, as they reduce
taxes on dividends and therefore make dividend payouts more attractive from
the point of view of shareholders.
There is another less speculative effect that operates through the cost of
capital. Lower taxes on dividend income may reduce the cost of capital for
investment financed by issuing new shares, since the anticipated return in the
form of future dividend payments is subject to lower taxation. In aggregate,
however, most investment is financed not by new equity but by either retained
earnings or borrowing.11 The cost of capital using these sources of finance will
not change. In particular, since financing investment by retained earnings
10
The other three countries are Switzerland, Luxembourg and the Netherlands. There are
other countries that do not have imputation systems; however, as they have reduced income
tax rates on dividends, they have in effect relieved some of the double taxation and are
therefore not classified here as pure classical systems. These countries include Austria,
Belgium, Hungary, Japan, Poland and Sweden.
11
Some firms depend more heavily on new equity finance – for example, small start-up firms.
70
Company taxation and innovation policy
exchanges dividends now for the expectation of higher dividends in the future,
a permanently lower tax rate on dividends has no effect on the required rate of
return. Even for firms using new equity finance, the cost of capital will only
fall if their key shareholders are taxpaying individuals, affected by the
proposed reform. A large proportion of company equity is owned indirectly
through pension funds or plans, which are already exempt from income tax on
dividends. The cost-of-capital effect could therefore lead to higher investment
for some firms as a result of the proposed US tax change, but it is doubtful that
the aggregate effect will be large.
The proposed US reform may have other effects. Some stimulus to the stock
market is possible, but again not likely to be large, as domestic taxpaying
individuals are the only shareholders for whom taxation will change as a result
of the reform. More likely effects are changes in the way companies pay out
cash to shareholders. The current US tax system taxes dividends more highly
than share buy-backs, and arguably this has led to high levels of buy-backs in
the USA. Removing the tax disadvantage of dividends may well lead to firms
switching from buy-backs to dividend payments. But it should be stressed that
this is a change to the form in which cash is paid from firms to their
shareholders, not to the level of cash distributed by firms, and consequently it
is unlikely to have any substantive effects.
To sum up, while many countries have recently changed their dividend tax
systems, it is unlikely that the effects on business investment or the stock
market have been large. Specifically in the case of the USA, it is difficult to
avoid the conclusion that the main effect of the proposed US reform would be
to raise the post-tax incomes of individuals who own equities directly rather
than through tax-exempt pension plans, with the largest beneficiaries likely to
be among the wealthiest stockholders. Whilst such a measure is unlikely to
appeal to Gordon Brown, further changes to dividend taxation in the UK
should not be ruled out. Whether the UK can keep its current system, which
still uses tax credits, will also depend on international developments, such as
rulings by the European Court of Justice and the possible drafting of further
European Union directives.
6.3 North Sea taxation
Changes to North Sea taxation announced in Budget 2002 have some merits,
but policy in this area still seems to be driven too much by short-term revenue
needs or changes in the oil price. More consideration should be given to
creating a stable tax regime that would facilitate long-term investment
decisions.
The changes announced in the April 2002 Budget include the introduction of a
new supplementary charge in addition to corporation tax and new 100%
capital investment allowances, plus the abolition of licence royalties. The first
two changes applied with immediate effect, but the abolition of licence
royalties was subject to consultation on its timing. In the November 2002 PreBudget Report, it was announced that they would be abolished from 1 January
2003.
71
Green Budget, January 2003
Box 6.1. North Sea taxation
The following taxes and charges are levied on North Sea production. Most
remaining taxes are charged on a measure of profits, but the exact definition
of taxable profits varies across taxes. The taxes charged vary with the date of
approval of a field. A summary is given in Table 6.1.
Petroleum revenue tax (PRT): Charged on profits at a rate of 50% on fields
approved prior to 15 March 1993.
Corporation tax (CT): Charged on profits at a rate of 30% on all fields. This
is the same tax as charged on the UK mainland, except that it is ring-fenced,
i.e. losses from other activities cannot be set against profits from oil and gas
production.
Supplementary charge: Charged on profits at a rate of 10% on all fields.
Unlike for corporation tax purposes, financing expenses (mainly interest) are
not deductible.
Prior to 1 January 2003, there were also:
Licence royalties: Charged on gross value of output, less some limited
expenditure, at a rate of 12.5% on fields approved prior to 31 March 1982.
Table 6.1. Applicable taxes and marginal tax rates by date of approval
of field
Date of approval
of oilfield
Up to 31/3/82
1/4/82–15/3/93
Since 16/3/93
Taxes that apply:
Marginal tax rate
CT, supplementary charge,
PRT, licence royalties (until
31/12/02)
CT, supplementary charge,
PRT
CT, supplementary charge
73.8% until 31/12/02
70% from 1/1/03
70%
40%
Economic effects of the reforms
The new 10% supplementary charge and the 100% capital allowances are in
line with the economic theory of resource taxation. Many economists have
argued that taxes levied on profits from the exploitation of natural resources
should be neutral with respect to investment. In other words, they should not
discourage investment in projects that would be profitable in the absence of
taxation. This can be achieved by taxing economic rents rather than total
profits. Economic rents are any profits in excess of the minimum level that
makes a project commercially viable. Extracting natural resources typically
generates a high level of economic rents, as the underlying resources are
intrinsically scarce. The level of rents an oilfield generates also depends
strongly on the price of crude oil and therefore varies if prices are volatile.
Capital allowances of 100% ensure that a large share of the minimum required
72
Company taxation and innovation policy
return to a project remains untaxed,12 while the supplementary charge
increases the tax rate on rents. These two changes together can thus be seen as
a step towards economically efficient rent taxation for new investment activity
in the North Sea.
The remaining change is the abolition of licence royalties. This only affects
oilfields approved prior to 31 March 1982, as only those fields were liable to
royalties. Royalties were levied on the gross value of output, less some limited
expenditure. Being revenue-based, royalties can have extremely harmful
effects on incentives to invest. Not only are they charged on fields earning less
than the minimum required return, but they can even apply to loss-making
fields, as most costs are not deductible in the calculation of their base. The use
of royalties also meant that there were three different tax regimes coexisting in
the North Sea, depending on the date of approval of a field. Their abolition
reduces this number to two. (See Box 6.1 for a summary of North Sea
taxation.)
Effects on revenues
The combined effect of the 10% supplementary charge and the new 100%
investment allowances is to increase tax revenues by about £600 million
yearly by 2004–05.13 Licence royalties raised £558 million in 2001–02,14 just
under 11% of total tax revenues from the North Sea. However, as royalties
were deductible as an expense for all other North Sea taxes, the revenue cost
of their abolition is estimated to average £143 million per year.15 The net
effect of all the new changes to North Sea taxes will therefore be to raise
annual revenues by nearly half a billion pounds.
Conclusion
Although the recent changes have merit in themselves, it is important to view
North Sea taxation in historical perspective. Too often in the past, North Sea
tax rules have changed in response to short-term revenue needs or changes in
the oil price. Investment in the North Sea needs to be planned over a long time
horizon, and stability is needed to allow firms to make informed decisions.
This implies the need for a stable tax regime that is able to cope automatically
with the volatility of oil prices.
The current reforms arguably did not go far enough as there are still two
different tax regimes, depending on the date of approval of a field. While
opinions on the optimal level of taxation may differ, it is hard to see how
12
Some normal profits will still be taxed, because the 100% allowance does not apply to all
capital investment and because the value of any unused allowances diminishes in presentvalue terms.
13
Table A.1 of HM Treasury, Financial Statement and Budget Report, London, 2002
(www.hm-treasury.gov.uk/Budget/bud_bud02/budget_report/bud_bud02_repchapa.cfm?).
14
Table 11.11 of Inland Revenue Statistics
(www.inlandrevenue.gov.uk/stats/corporate_tax/ct_t11_1.htm).
15
Table B.4 of HM Treasury, Pre-Budget Report 2002, Cm. 5664, London, 2002 (www.hmtreasury.gov.uk/Pre_Budget_Report/prebud_pbr02/report/prebud_pbr02_repannexb1.cfm?).
73
Green Budget, January 2003
different rates on fields approved at different times form part of a welldesigned tax system. The argument that high taxes on old fields do not cause
any harm because investment is a sunk cost does not completely hold. As
older fields reach the end of their lifetimes, incremental investment is often
needed. Premature abandonment of fields would be undesirable, as the cost of
reopening fields later is likely to be prohibitive and any unused resources
would potentially be lost.
6.4 Stamp duty on shares
In 2001–02, stamp duty on share transactions raised £2.9 billion, down from
£4.5 billion in 2000–01. This represents about ¾% of total public sector net
receipts, a small but not insignificant sum.16 The Chancellor’s fiscal room for
manoeuvre is clearly limited in the forthcoming Budget, but given his oftstated goal of increasing productivity, there is no reason why, in principle, he
could not consider policy change involving a reduction in the rate of stamp
duty on shares.
Stamp duty is levied at ½% of the purchase price on all share transactions in
UK incorporated companies. Reducing stamp duty on shares – perhaps instead
of a reduction in a less distorting tax such as corporation tax – would be likely
to boost productivity for three main reasons. First, stamp duty lacks any
investment allowances and is therefore likely to discourage investment more
than other kinds of capital taxes. Secondly, it reduces the efficiency of the
stock market for UK listed companies by raising transactions costs and
possibly increasing share price volatility. Finally, it distorts merger and
acquisition activity, producing a bias towards overseas rather than UK
ownership of companies. Each of these effects is discussed in turn below.
Stamp duty is levied on the purchase price of a share. Unlike corporation tax,
which taxes profits after allowing at least partially for the cost of investment,
stamp duty in effect taxes both the full amount invested and the subsequent
profits. Thus, for a given revenue yield, stamp duty imposes a heavier tax
burden on investments that just break even, making it more likely that the tax
will prevent them taking place. Stamp duty on shares is therefore likely to be a
less efficient way of raising revenue than corporation tax.
Because stamp duty is levied on transactions, it directly reduces share turnover
and market liquidity, thereby reducing the efficiency of the market in UK
company shares wherever they are traded. The limited empirical evidence
available suggests that reducing the rate of stamp duty on shares may increase
share turnover substantially.17 Transactions taxes such as the ‘Tobin tax’ on
foreign exchange transactions have been proposed as a way of reducing price
volatility and therefore risk in financial markets by discouraging short-term
speculative behaviour. However, stamp duty may actually increase volatility
16
Inland Revenue Statistics (www.inlandrevenue.gov.uk/stats/index.htm).
17
See table 3.1 of M. Hawkins and J. McCrae, Stamp Duty on Share Transactions: Is There a
Case for Change?, Commentary no. 89, IFS, London, 2002
(www.ifs.org.uk/corptax/comm89.pdf).
74
Company taxation and innovation policy
by reducing the liquidity of the market and increasing the price impact of a
given size of share transaction. The empirical evidence is mixed, but on
balance it does not support the idea that stamp duty on share transactions is
likely to reduce price volatility.18
Stamp duty may also distort the market for corporate control. A UK company
planning a foreign takeover will be willing to pay less than an otherwise
identical overseas competitor for the target company, due to the stamp duty
that its shareholders will have to pay on future share transactions in the foreign
subsidiary. On the other hand, a foreign company considering a takeover of a
UK company will be willing to pay more for it than otherwise identical UK
companies, since it will take the shares in the company outside the stamp duty
base. The resulting distortion to merger and acquisition activity means that
companies may end up not being run by the set of managers that would deliver
the best performance.
In the last Budget, which raised taxes overall, the Chancellor still found room
to spend £900 million on measures designed to raise productivity.19 If he
wishes to focus on this goal again, then taking together the effects described
above, there may be a case for reducing stamp duty on shares to alleviate these
distortions.
6.5 R&D and policy towards innovation
Following the introduction of a research and development (R&D) tax credit
for larger firms in Budget 2002, the 2002 Pre-Budget Report went on to
announce an independent review into collaboration between business and
universities. The review will report by Summer 2003, around the same time as
the Department of Trade and Industry (DTI) Innovation Review, which is to
examine business innovation and its contribution to UK productivity growth.
This section documents trends in UK R&D over the last two decades and
discusses the role of government in supporting R&D and innovation. It
summarises recent changes to the tax treatment of R&D, and discusses the
current direction of government policy towards innovation.
Trends in UK R&D
Gross expenditure on R&D as a percentage of GDP or national income
(GERD intensity) has declined steadily in the UK over the last two decades,
while in other G5 countries it has either increased or shown little overall
change. In particular, while all of the G5 experienced stagnant or falling
GERD intensity between 1990 and 1994, the USA, Japan and Germany all
showed strong increases over the second half of the 1990s. In contrast, the UK
18
See section A.3 of M. Hawkins and J. McCrae, Stamp Duty on Share Transactions: Is There
a Case for Change?, Commentary no. 89, IFS, London, 2002
(www.ifs.org.uk/corptax/comm89.pdf).
19
HM Treasury, Budget 2002: The Strength to Make Long-Term Decisions, London, 2002
(www.hm-treasury.gov.uk/Budget/bud_bud02/bud_bud02_index.cfm?).
75
Green Budget, January 2003
continued to decline after 1994, with signs of a small pick-up only emerging
after 1998. The result is that the UK now has levels of GERD intensity up to 1
percentage point of national income lower than other G5 countries. (See
Figure 6.1.)
Figure 6.1. GERD as a percentage of GDP: G5 countries
GERD as a % of GDP
3.0
2.5
2.0
1.5
1981
1984
1987
1990
1993
1996
1999
Year
UK
USA
France
Germany
Japan
Note: Data for Germany cover West Germany until 1990 and unified Germany from 1991.
Source: OECD, Main Science and Technology Indicators, Paris, 2002 (www.sourceoecd.org).
GERD is composed of three main constituent parts according to who performs
the R&D. By far the largest component is business enterprise expenditure on
R&D (BERD), followed by higher education expenditure on R&D (HERD)
and government expenditure on R&D (GOVERD). GERD can also be broken
down according to who funds the R&D, which may not be the same as who
performs it. The main sources of funding are government (including Research
Councils and Higher Education Funding Councils), domestic business
enterprise, abroad, and other national sources (mainly private non-profit
organisations). Breakdowns of GERD both by who performs the R&D and by
who funds it are shown in Table 6.2.
The shares of GERD funded and carried out by government fell sharply
between 1981 and 1990, with a smaller decline between 1990 and 2000.20 The
share funded from abroad has been rising steadily over the period. Within
BERD, there has also been a shift in funding away from government and
towards domestic business and overseas sources, as Table 6.3 shows.
20
The Atomic Energy Authority was transferred from the government sector to the business
enterprise sector after 1986. While the effect of this transfer and later privatisations is to
overstate slightly the extent of the trend away from government activity, the overall picture is
not significantly affected.
76
Company taxation and innovation policy
Table 6.2. Breakdown of R&D by who performs it and who funds it
GERD as a % of GDP
GERD, £ma
% of GERD by who performs it:
BERD
HERD
GOVERD
Of which: defence
civil
% of GERD by who funds it:
Government
Of which: defence
civil
Domestic business
Of which: defence
civil
Abroad
Other national sources
1981
2.38%
13,720
1990
2.15%
16,381
2000
1.86%
17,532
63%
14%
21%
–
–
71%
16%
13%
6%
7%
66%
21%
12%
8%
4%
50%
–
–
41%
–
–
7%
2%
34%
14%
20%
50%
3%
47%
12%
4%
29%
8%
21%
49%
3%
46%
16%
6%
a
In millions of 2001 pounds sterling, deflated by the GDP deflator.
Note: Figures do not sum to 100 as the private non-profit sector has been omitted from the
table.
Sources: OECD, Main Science and Technology Indicators, Paris, 2002
(www.sourceoecd.org); ONS, Gross Expenditure on Research and Development, London,
2002 (www.statistics.gov.uk).
Table 6.3. Percentages of BERD by who funds it
% of BERD by source of funding
Government
Domestic business
Abroad
Other national sources
1981
30%
61%
9%
–
1990
16%
68%
16%
–
2000
9%
70%
21%
–
Source: OECD, Main Science and Technology Indicators, Paris, 2002 (www.sourceoecd.org).
Almost all of the fall in GERD intensity over the 1990s is due to a fall in
BERD intensity, its largest component. In particular, Figure 6.2 shows that
BERD intensity continued to fall from 1994 to 1998 in the UK, while it was
rising in the USA, Japan and Germany, and only began to recover after 1998.
In fact, the level of UK spending on BERD was only very slightly higher in
real terms in 2000 than in 1990. BERD is particularly significant because it is
the most commercially relevant component of GERD, and it is also the
component that should be affected by the new R&D tax credits.
A key feature of UK R&D performance over the 1990s has been the rapidly
increasing amount of R&D done by UK firms abroad, especially in the USA.
Figure 6.3 shows total levels of UK R&D from two different sources over the
period 1992–2000. The first is BERD as discussed above, which includes all
R&D that is performed in the UK, while the second comes from the DTI’s
R&D Scoreboard. This lists all R&D done by UK firms including their
77
Green Budget, January 2003
subsidiaries abroad, plus R&D done by UK subsidiaries of foreign firms, as
reported in company accounts. The R&D Scoreboard figure thus corresponds
Figure 6.2. BERD as a percentage of GDP: G5 countries
BERD as a % of GDP
2.5
2.0
1.5
1.0
1981
1984
1987
1990
1993
1996
1999
Year
UK
USA
France
Germany
Japan
Notes: Data for the UK and Germany extend up to 2001. Data for Germany cover West
Germany until 1990 and unified Germany from 1991.
Source: OECD, Main Science and Technology Indicators, Paris, 2002 (www.sourceoecd.org).
Figure 6.3. UK R&D levels: BERD and R&D Scoreboard
R&D level, 1992=100
200
150
100
50
0
1992
1994
1996
1998
2000
Year
BERD
R&D Scoreboard
Note: Nominal levels have been deflated by the GDP deflator and rebased to equal 100 in
1992.
Sources: OECD, Main Science and Technology Indicators, Paris, 2002
(www.sourceoecd.org); Department of Trade and Industry, R&D Scoreboard, 1993–2001.
78
Company taxation and innovation policy
roughly to the BERD figure plus the R&D done by UK firms abroad.21 The
graph shows that R&D spending by UK firms has grown much faster over the
1990s than R&D performed in the UK. This discrepancy raises the question of
whether we should be concerned from a policy point of view with R&D that is
located in the UK or R&D that is done by UK firms. The latter may be more
important if firms are locating R&D abroad in order to source new
technologies from the cutting edge of innovation.22
Figure 6.4. UK BERD as a percentage of GDP: industry breakdown
BERD as a % of GDP
1.5
Services
1.0
Aerospace
Motor vehicles
Pharmaceuticals
0.5
Chemicals
Machinery and equipment
Other manufacturing
0.0
1987
1990
1993
1996
1999
Year
Source: OECD, Analytical Business Enterprise Research and Development, Paris, 2002
(www.sourceoecd.org).
The industry composition of UK BERD has changed over the last two
decades. Figure 6.4 shows UK BERD intensity between 1987 and 2000
broken down into broad industrial sectors. Most R&D is done in
manufacturing industries, with ‘pharmaceuticals’ now the largest single
contributor to aggregate BERD. Most of the decline in overall intensity during
the 1990s can be explained by the ‘chemicals’, ‘machinery and equipment’
and ‘other manufacturing’ sectors. ‘Pharmaceuticals’ is the only sector to have
increased its contribution to BERD intensity over the period, replacing
‘machinery and equipment’ as the largest contributor in about 1996.
A fall in aggregate BERD intensity over time can be decomposed into two
contributing factors: a fall in intensity within industries, and a shift in the
21
There are some differences in the definition of eligible R&D between the two. For example,
the R&D Scoreboard does not include R&D done under contract for government or other
firms. These differences should not seriously affect the comparison of overall trends.
22
See M. Serapio and D. Dalton, ‘Globalization of industrial R&D: an examination of foreign
direct investments in R&D in the United States’, Research Policy, 1999, vol. 28, pp. 303–16.
The authors suggest that this ‘technology-sourcing’ behaviour is increasingly common
amongst foreign firms that locate R&D activity in the USA.
79
Green Budget, January 2003
composition of output away from high-intensity towards low-intensity
industries. Figure 6.4 does not distinguish between declines in intensity within
industries and shifts in activity between industries. Almost all OECD countries
have seen their service sectors grow faster than manufacturing over the last
twenty years, and this would automatically tend to reduce aggregate BERD
intensity due to the second, between-industry factor. However, almost all of
the decline in aggregate BERD intensity in the UK relative to other G5
countries during the mid-1990s is due to a fall in R&D intensity within
manufacturing industries, rather than to a particularly rapid decline in the
share of manufacturing activity in the UK over this period.23
Government policy towards innovation
The government intervenes in innovation markets in a number of ways. As
shown above, a significant part of R&D expenditure is financed directly by
government. Other forms of intervention include R&D tax credits, support for
technology transfer between universities and business, and measures designed
to overcome failures in financial markets. In addition, government funds
education and other aspects of national infrastructure that are essential inputs
into R&D and innovation, and provides patent protection and regulation. In
this section, we briefly describe some of these measures and discuss the
rationales behind them.
Direct support
Direct government support for innovation largely consists of funding for
R&D. This funding comes through one of three channels: directly from central
government departments (about 50%), from the Research Councils (about
25%) and from the Higher Education Funding Councils (about 25%).24 Where
the R&D is actually carried out varies between the different channels. For
example, in 2000, about 45% of R&D funded directly by government
departments was performed in government laboratories and about 40% was
performed by businesses, with the majority of the rest performed in higher
education institutions. In contrast, almost all of the R&D funded through the
other two channels was carried out in higher education institutions.
There are several reasons why government may want to fund or carry out
R&D. In the case of defence R&D, the government is a provider of a service –
namely, national defence – to the country. The government has considerable
information advantages over the private sector as regards its defence needs,
suggesting a role for government in funding defence-related R&D. Whether
the government carries out this type of R&D itself or funds the private sector
to carry it out should be decided on grounds of cost-efficiency and perhaps
also the need for secrecy. In 2000, just under a third of all R&D funded by
government was for defence purposes. This contrasts with only 6% of R&D
23
See R. Griffith and R. Harrison, IFS Working Paper, forthcoming.
24
ONS, R&D Performed in the UK in Each Sector According to Source of Finance, London,
2002 (www.statistics.gov.uk/STATBASE/tsdataset.asp?vlnk=584).
80
Company taxation and innovation policy
funded by businesses, although over half of the defence R&D funded by
government was actually carried out by businesses.25
Another reason why the government may want to fund R&D is that it would
not otherwise be funded by the private sector because it would not be
profitable for any particular firm to fund it. R&D covers a spectrum, with
general scientific research that does not have a specific commercial use in
mind at one end, and R&D designed around the introduction of a specific
commercial product or process at the other. The market will provide weaker
incentives to undertake the former kind of R&D because it is more difficult for
an individual firm to appropriate the benefits of this type of research. In this
context, new knowledge can be thought of as a public good, in that once it is
generated it can be used by everyone. Thus, in the absence of government
intervention, if all R&D were funded by businesses, there would be too little
of this type of research from society’s point of view.26
These kinds of concerns apply particularly to fundamental scientific research.
This is the main economic argument behind government funding for research
in the higher education sector, via the Research Councils and Higher
Education Funding Councils. These institutions should ideally act as agents of
government to commission and deliver research that fulfils the needs of
society as a whole and that would not be provided by the private sector.
As mentioned above, government also directly funds R&D performed by
businesses. This direct funding in the UK has declined in real terms over the
last twenty years. Over the last decade, about 75% of it has been for defence
purposes, with the remainder for civil R&D. Again, the idea that businesses
cannot completely capture all the benefits of their research forms the main
justification for this kind of funding.
R&D tax credits
Direct funding is not the only way that the government can support R&D
performed by the private sector. In recent years, there has been a general trend
in many OECD countries away from direct grants and towards indirect
assistance via the favourable tax treatment of R&D expenditure. One reason is
that businesses may, in many cases, have better information than government
as regards which programmes of research are likely to be successful. Tax
credits attempt to stimulate private sector R&D by reducing its cost while
keeping control over the nature and direction of research in the hands of
businesses. With the introduction of the R&D tax credit for small and
medium-sized enterprises (SMEs) in April 2000 and for larger firms in April
2002, the UK has now joined several other OECD countries in supporting
private sector R&D by this method.
The SMEs tax credit rate is 50% and the rate for larger firms is 25%. Both
credits operate as an extra deduction, which in practice means that qualifying
25
ONS, Sources of Funds for Civil and Defence R&D in UK Businesses, London, 2002,
www.statistics.gov.uk/STATBASE/tsdataset.asp?vlnk=571.
26
More generally, the patent system is one way in which government intervenes to overcome
this appropriation problem: ensuring that inventors are able to profit from their inventions
protects their incentives to innovate.
81
Green Budget, January 2003
SMEs are able to deduct 150%, and larger firms 125%, of eligible current
R&D expenditure from their taxable profits in the year it is incurred.27 The
SMEs credit also has a refundable component. Qualifying SMEs with
insufficient taxable profits to claim the full deduction can claim a cash
payment equal to 24% of eligible expenditure, instead of carrying forward
eligible R&D losses to offset against future profits. The amount payable is
limited to the company’s PAYE and National Insurance contributions for the
period.
The SMEs credit was projected to cost the exchequer around £100 million in
lost revenue in 2001–02 and £150 million per year from 2002–03, potentially
benefiting around 4,500 firms.28 The actual cost in 2001–02 is estimated in the
2002 Pre-Budget Report to have been only £80 million,29 although it is as yet
unclear whether this was due to lower-than-expected take-up or lower R&D
per SME. The equivalent cost of the tax credit for large firms was forecast to
be £200 million in the first year, rising to £400 million by 2004–05.30
Estimates of actual costs will not be available until after the end of the current
tax year. These figures compare with £1.5 billion of direct government
funding for private sector R&D in 2001, of which only £190 million was for
non-defence purposes.31
The 2002 Pre-Budget Report states that one of the areas to be covered by the
forthcoming review into business–university collaboration will be the
effectiveness of the R&D tax credits in stimulating business demand for
research and skills. While evaluation is extremely important for the design of
successful policy, it should be stressed that the effects of these policies are
only likely to be fully realised over a longer time period. Evidence from the
USA and other countries suggests that firms’ responses to the introduction of
R&D tax credits are characterised by long lags while investment and research
plans respond to the new incentives, with the full response only realised after
as long as 10 years.32 This suggests that the government should resist the
temptation to alter policies significantly in the short term, so that the policy
regime has a chance to ‘bed down’.
The interaction between the R&D tax credits and other forms of government
support for innovation differs between the two tax credits. The R&D tax credit
27
The R&D allowance allows firms to deduct 100% of capital expenditure on R&D from their
taxable profits. This is already more generous than standard capital allowances.
28
HM Treasury, Financial Statement and Budget Report, London, 1999
(http://archive.treasury.gov.uk/budget/1999/fsbr/29807.htm).
29
HM Treasury, Pre-Budget Report 2002, Cm. 5664, London, 2002 (www.hmtreasury.gov.uk/pre_budget_report/prebud_pbr02/prebud_pbr02_index.cfm).
30
Table A.1 of HM Treasury, Pre-Budget Report 2002, Cm. 5664, London, 2002 (www.hmtreasury.gov.uk/budget/bud_bud02/budget_report/bud_bud02_repchapa.cfm).
31
ONS, Sources of Funds for Civil and Defence R&D in UK Businesses, London, 2002
(www.statistics.gov.uk/STATBASE/tsdataset.asp?vlnk=571).
32
See, for example, N. Bloom, R. Griffith and J. Van Reenen, ‘Do R&D tax credits work?
Evidence from an international panel of countries, 1979–1994’, Journal of Public Economics,
2002, vol. 85, pp. 1–31.
82
Company taxation and innovation policy
for SMEs is unavailable on projects that have benefited from notified State
Aids such as DTI Smart or Link awards, and is restricted to the unsubsidised
portion of R&D expenditure for projects benefiting from non-State Aids.33
Larger firms, on the other hand, can claim the R&D tax credit even on R&D
that has been partly or fully subsidised directly. They also receive the full
credit on R&D subcontracted to universities and other non-profit bodies such
as scientific research organisations or NHS bodies, but not on other
subcontracted R&D. SMEs receive a credit on 65% of all subcontracted R&D
expenditure provided they retain intellectual property rights.
Other policies
Government policy towards science and innovation has also focused on the
science base and on the efficient use of research – for example, through
knowledge transfer from universities to businesses. In July 2002, the
government published its science strategy, Investing in Innovation,34 and the
November Pre-Budget Report announced an independent review into
business–university collaboration. The review will report by Summer 2003,
around the same time as the DTI Innovation Review, which is to examine
business innovation and its contribution to UK productivity growth.
There are currently a large number of relatively small-scale policies aimed at
enhancing knowledge transfer, especially between universities and businesses.
These include the Higher Education Innovation Fund, with funding projected
to be £90 million by 2005–06, and other DTI knowledge-transfer activities,
projected to cost £300 million by 2005–06. Together with the Wellcome Trust
and the Gatsby Foundation, the government has funded the University
Challenge scheme. This is aimed at encouraging commercial spin-offs from
university research, with a total £30 million of initial investment capital so far
combined with £40 million from private sector sources.35
The forthcoming reviews might provide a good opportunity to examine the
economic rationales for such schemes, the incentives each scheme provides
for innovation and commercialisation of research, and the interactions between
different policies. Economic justification of such policies should identify
specific market failures – for example, those leading to insufficient uptake by
businesses of new technologies developed in universities; in the case of the
University Challenge scheme, a case might be made that private investors do
not have sufficient information to determine whether to invest in new
technologies. A further issue is whether any of the schemes distort firms’
decisions about how to implement and organise new innovations. For
example, the University Challenge scheme only funds the commercialisation
33
The legal definition of State Aids is set out in Article 87(1) of the Treaty of Rome. A
subsidy is considered a State Aid when the effect of aid: distorts competition; is selective in its
effects (e.g. only affects subgroups of firms, or only affects businesses in a specific region or
locality); or affects trade, or could potentially affect trade, between EU member states.
34
HM Treasury, Department of Trade and Industry and Department for Education and Skills,
Investing in Innovation: A Strategy for Science, Engineering and Technology, London, 2002
(www.hm-treasury.gov.uk/spending_review/spend_sr02/spend_sr02_science.cfm).
35
HM Treasury, Pre-Budget Report 2002, Cm. 5664, London, 2002 (www.hmtreasury.gov.uk/pre_budget_report/prebud_pbr02/prebud_pbr02_index.cfm).
83
Green Budget, January 2003
of research through start-ups and not through licensing agreements, which
may be more suitable in some cases.
Stephen Bond, Rupert Harrison, Mike Hawkins,
Alexander Klemm and Helen Simpson
84
7. Childcare subsidies
The link between the affordability of good-quality childcare and the ability of
mothers to undertake paid employment remains at the centre of policy debate,
even though a higher proportion of mothers are currently active in the formal
labour market than ever before. There are concerns that mothers are
constrained from paid employment by a lack of suitable childcare options1 and
that care costs are rising, consuming a greater proportion of families’ financial
resources. In addition, there is an ongoing debate about whether the type of
care that is available and affordable is beneficial to children.
In his Pre-Budget Report of November 2002, the Chancellor said the
government wanted to help ‘parents to make real and effective choices on
balancing work and family life’. Over the last decade, several initiatives have
been introduced to help families with the cost of childcare. A childcare
deduction was introduced in family credit in 1994 and expanded with the
introduction of the working families’ tax credit (WFTC) in 1999. The
Childcare Voucher Scheme (later the nursery education grant) was
implemented in 1997. In April, the childcare credit currently included in the
WFTC will transfer to the new working tax credit (WTC). The types of care
eligible for the childcare credit will be extended in a limited way ‘to include
those who use approved childcare in their own home, benefiting, among
others, parents of disabled children and those who work outside conventional
hours’ (italics added).2 Further plans for the expansion of the types of care
covered by the credit have been announced, although many are relatively
minor extensions.3
This chapter considers the potential effects on the number of eligible families
and the annual budget cost of further expanding the scope and generosity of
the childcare credit. The estimates are initially made under the unrealistic
1
For example, see paragraph 3.26 on page 21 of HM Treasury and Department of Trade and
Industry, Balancing Work and Family Life: Enhancing Choice and Support for Parents,
January 2003 (www.dti.gov.uk/er/individual/balancing.pdf).
2
Page 91 (chapter 5) of HM Treasury, Pre-Budget Report 2002, Cm. 5664, London, 2002
(www.hmtreasury.gov.uk/pre_budget_report/prebud_pbr02/report/prebud_pbr02_repchap5.cfm)
3
It has been announced that the Home Childcarers Scheme will be opened to existing
childminders and that consideration will be given to ‘how to widen entry into the scheme to
include people who are not already childminders’. The type of care eligible for the childcare
credit will also be broadened by expanding the regulation required for eligibility to
domiciliary workers and nurses employed through agencies and to providers who offer
childcare exclusively to children over the age of 7 (paragraphs 4.59 and 4.60 on page 36 of
HM Treasury and Department of Trade and Industry, Balancing Work and Family Life:
Enhancing Choice and Support for Parents, January 2003
(www.dti.gov.uk/er/individual/balancing.pdf)). In addition, the government is considering
reforms to tax and National Insurance incentives to expand employer-supported childcare (box
1.1 on page 3 of HM Treasury and Department of Trade and Industry, Balancing Work and
Family Life: Enhancing Choice and Support for Parents, January 2003
(www.dti.gov.uk/er/individual/balancing.pdf)).
85
Green Budget, January 2003
assumption that childcare and employment behaviour remains unaltered by the
changes, but this is followed by a discussion of the possible reactions in family
childcare and work choices.4
Section 7.1 briefly reviews the arguments for government subsidies in the
childcare market, while Section 7.2 describes the current employment and
childcare choices of families. A description of options for expanding the
childcare credit is presented in Section 7.3, followed by the estimated potential
impacts in Sections 7.4–7.7. Section 7.8 concludes.
7.1 Why subsidise childcare?
The case for subsidising childcare is essentially twofold. First, it is argued that
mothers should be encouraged to work in formal paid employment: to reduce
gender inequalities in the labour market; to make the best use of the potential
labour force and thereby improve economic efficiency; and to reduce the
dependence of poorer households on the state. Secondly, if pre-school children
benefit from childcare (other than that of family), it may be desirable on equity
grounds to ensure that it is available to poorer families.
In the absence of government intervention, families may not make the best
employment and childcare choices, for several reasons: the benefits may be
social as well as private; parents may lack complete information or be shortsighted; or parents may be unable to afford adequate childcare. For these
reasons, it may be desirable for government policies to help families with the
cost of care.5 The design of the subsidy may depend upon the particular
rationale and potential cost.
4
Budget costs of different childcare subsidies have been estimated previously in A. Duncan,
C. Giles and S. Webb, The Impact of Subsidising Childcare, Research Discussion Series
no.13, Equal Opportunities Commission, Manchester, 1995, and in A. Duncan and C. Giles,
‘Should we subsidise pre-school childcare, and if so, how?’, Fiscal Studies, 1996, vol. 17, no.
3, pp. 39–61. These publications use data from the Family Resources Survey from the early
1990s together with childcare information from the 1991–92 General Household Survey. They
include potential labour supply responses, but only consider formal types of childcare for preschool children and limit changes in childcare choices to those related to employment
responses. The budget cost of a particular childcare tax credit has also been analysed using
aggregate statistics in Ernst & Young, ‘Potential cost of the childcare tax credit’, Tax Policy
Discussion Paper, 2001.
5
In addition, there may be supply constraints in the provision of childcare due to
inefficiencies in the childcare market. This problem may be addressed by more direct
measures to improve the supply of childcare, such as those included in the government’s
National Childcare Strategy and financed by the combined budgets for Sure Start, childcare
and early years programmes (paragraph 2.2 on page 1 of HM Treasury and Department of
Trade and Industry, Balancing Work and Family Life: Enhancing Choice and Support for
Parents, January 2003 (www.dti.gov.uk/er/individual/balancing.pdf)).
86
Childcare subsidies
7.2 Families’ employment and
childcare use
This section outlines the current employment status of parents and the patterns
of childcare use by working families. These tables and the analysis below use
a sample of families with dependent children (those aged under 16, or under
19 and in full-time education) from three years of the Family Resources
Survey (1998–99 to 2000–01).6
Over three-quarters of families (78%) have at least one parent working in paid
employment for 16 or more hours each week. Some 39% of all families have
two working adults, while single parents working 16 or more hours each week
constitute 10% of all families. But, as Table 7.1 shows, the likelihood that a
parent is in paid employment depends considerably on whether there is a
partner present and on the age and number of children in the family. Single
parents (predominantly single mothers) are much less likely to be working
than mothers with a partner. Mothers are less likely to be in employment if at
least one child is of pre-school age or if there is more than one child.
Table 7.1. Employment patterns among families with children
Percentage of families with
weekly hours of work:
no one works 16+
only mother works 16+
only father works 16+
both parents work 16+
single parent works 16+
Couples
Pre-school
Only school
children
children
1
2+
1
2+
Single parents
Pre-school
Only school
children
children
1
2+
1
2+
All
7
2
38
53
–
73
–
–
–
27
22
3
26
39
10
10
2
53
35
–
9
5
23
62
–
9
4
31
56
–
83
–
–
–
17
48
–
–
–
52
59
–
–
–
41
Number of families (1,000s)
679 1,207 1,344 1,977 255
316
615
559 6,952
Notes: The category ‘pre-school children’ includes families with at least one pre-school child.
The symbol ‘–’ denotes a cell that is not applicable to the family group. The numbers of
families have been grossed to the national level.
Source: Family Resources Survey (1998–99, 1999–2000 and 2000–01)
To be eligible for the childcare credit in the current working families’ tax
credit (and also for the childcare credit to be introduced in April 2003), both
parents in a couple and single parents must be in employment for at least 16
hours each week (with certain exceptions for incapacitated parents). Table 7.2
presents the current use of childcare for families fulfilling this requirement.
When asked whether ‘anyone else normally looks after the children because
you or your partner are working’, 44% of working families (defined as those
where all parents are in employment at least 16 hours each week) respond that
6
A full description of the survey and extensive analysis of the childcare information can be
found in G. Paull and J. Taylor with A. Duncan, Mothers’ Employment and Childcare Use in
Britain, Institute for Fiscal Studies, London, 2002.
87
Green Budget, January 2003
they use childcare and only 23% report that they pay for the care. Given that
almost three-quarters (73%) of the families with no parent working less than
16 hours each week contain only school-age children, this is not surprising as
parents may be able to fit work around school hours or the children may be old
enough to look after themselves. Indeed, Table 7.2 shows that the majority of
working families with pre-school children do use childcare, while around half
pay for it.
Table 7.2. Childcare use for families with no parent working less than 16
hours each week
Any type of non-parental care:
% using care
% paying for care
Average weekly cost
Formal care:
% using care
% paying for care
Average weekly cost
Informal care:
% using care
% paying for care
Average weekly cost
Couples
Pre-school
Only school
children
children
1
2+
1
2+
Single parents
Pre-school
Only school
children
children
1
2+
1
2+
All
77
50
£76
72
49
£90
23
10
£33
34
15
£42
85
51
£63
90
51
£68
38
16
£33
47
20
£37
44
23
£62
42
41
£80
40
39
£97
7
7
£35
10
10
£46
43
40
£68
39
37
£71
11
10
£35
12
11
£44
18
17
£69
47
19
£52
45
19
£49
18
5
£28
26
8
£33
56
24
£46
68
25
£48
31
9
£28
39
12
£25
31
11
£38
Number of families (1,000s)
361
419
830 1,112
70
54
320
229
3,394
Notes: The category ‘pre-school children’ includes families with at least one pre-school child.
The average weekly cost is the total cost for all children and is averaged over those paying for
care. All monetary values are in approximate April 2003 prices. Formal care includes
nurseries, crèches, school clubs, childminders, nannies and au pairs. Informal care is that
provided by relatives and friends. The numbers of families have been grossed to the national
level.
Source: Family Resources Survey (1998–99, 1999–2000 and 2000–01).
Childcare can be divided into two types: formal care, provided in a market
setting by nurseries, crèches, school clubs, childminders, nannies and au pairs;
and informal care, provided by relatives and friends. Childcare subsidies have
typically covered only the costs of formal care. Even so, only 18% of the
working families report using formal care, while 31% have used informal care
(the use of each type is not mutually exclusive, so the proportions sum to
slightly more than the total for ‘any type of childcare’). Almost all formal
care, but only about one-third of informal care, is paid for. The average
weekly cost (for those who pay) is higher for formal than for informal care.
Hence, families using formal childcare tend to spend more than users of
informal care, both because they are more likely to pay for it and because, if
they do pay, they pay more.
Working single parents use informal care more than working couples with
children. Families with only school children are also more likely to rely on
informal sources of care than families with at least one pre-school child.
Families with more than one child tend to pay less per child for care than those
with a lone child.
88
Childcare subsidies
These patterns in childcare use and the differences across different types of
families will have important implications for the effects of various kinds of
childcare subsidies. In particular, the substantial proportions of working
families not using care or using only informal care suggest that there might be
room for considerable expansion in the use of childcare by these families.
7.3 Options for subsidising childcare
This section reviews the current system of support for the childcare costs of
working families and how this will change from April 2003. It then describes
some modifications to the childcare credit that would increase the generosity
of the childcare subsidy.
Current programmes: the working families’ tax credit
The current childcare subsidy available in the working families’ tax credit has
the following features:
•
The subsidy is 70% of actual childcare expenditure, subject to a fixed
upper limit (a maximum credit of £94.50 for £135 of expenditure each
week for families with one child and £140 for an expenditure of £200 for
families with more than one child).
•
Only approved formal care is covered – namely: registered childminders,
play schemes or out-of-school clubs, other childcare schemes that do not
need to be registered and providers of care that are approved by
specifically accredited organisations.
•
To receive the care subsidy, all parents in the family are required to be in
paid employment at least 16 hours each week (with some exceptions for
incapacitated parents).
•
The subsidy is subject to a means test based on family net earnings.
Working tax and childcare credits to be introduced in April
2003
The working tax credit and child tax credit, which replace the WFTC and
children’s tax credit in April 2003, will basically replicate the current system
for families with children.7 There will be some adjustments:
•
The means test will be based on gross family income rather than net
earnings.
•
Some of the rates will be slightly adjusted.
•
The 30-hour premium in the WTC will be given to families where the
combined work hours of all parents are at least 30 each week (rather than
one parent being required to work at least 30 hours each week as in the
WFTC).
7
See Chapter 4 for further details on the new tax credits.
89
Green Budget, January 2003
•
The credits will be divided into a working tax credit (paid through the
employer), a childcare credit and a child tax credit (both paid directly to
the ‘prime carer’).
•
The childcare credit will benefit some families further up the income
distribution.
In considering alternative ways to subsidise childcare, the baseline WTC
scheme is subjected to several variations in three main dimensions.
Changing the amount of childcare expenditure covered by the
subsidy
Under the WTC, 70% of expenditure on formal types of childcare will be
refunded by the credit, up to ceilings of £94.50 (for £135 of expenditure) each
week for families with one child and £140 (for £200 of expenditure) for
families with more than one child. We consider two options that would allow
the amount of expenditure covered to be expanded:
1) Doubling the ceilings for the childcare credit to £189 for families with one
child and £280 for families with more than one child
2) Raising the eligible proportion of childcare costs to 100 percent
Options that extend the type of care covered by the credit are also considered.
Expanding coverage to informal care has been the subject of much discussion.
The Interdepartmental Childcare Review of November 2002 recognised the
importance of informal care for working families but expressed concern that
more evidence was required on ‘the outcomes of informal care for the child
and for parents using it; the extent to which the cost of informal care is a
barrier to parental employment; and whether paying informal carers would
lead to an overall growth in available childcare’8 before it could be decided
whether the government should intervene in the informal market. It might also
be desirable to know the potential cost. So the third option considered is:
3) Including informal care, such as that provided by friends and relatives, as
eligible for the credit
One difficulty with allowing the credit to cover informal care is the question
of how the price of care might be set. In particular, the 30% of the cost not
covered by the subsidy may not be a constraint on the price paid by the parent
if the money is going to a friend or relative. One way to address this would be
to set a maximum hourly rate. This would be equivalent to issuing vouchers
that could be cashed by the carer for each hour provided. Two variants of a
fourth option are therefore considered:
4a) A voucher scheme of £1.80 per child for each hour of care (formal or
informal), with a maximum number of eligible care hours equal to the
prime carer’s work hours (the prime carer in a couple being the parent
working the shortest hours, which is typically the mother)
8
Page 25 of Department for Education and Skills, Department for Work and Pensions, HM
Treasury and Women & Equality Unit, Delivering for Children and Families, London, 2002
(www.strategy.gov.uk/2002/childcare/report/downloads/su-children.pdf).
90
Childcare subsidies
4b) As for (4a) with a voucher value of £2.67 per child per hour
The figure of £1.80 is the average hourly amount spent on paid informal
sources of care, while £2.67 is the average hourly amount spent on care for
children in formal childcare. The voucher value could, in practice, be any
desired level of subsidy. It should be noted that the maximum value of the
subsidy under the voucher scheme depends not only on the number of children
in the family, but also on the work hours of the prime carer.
Changing the work requirements
Eligibility for the WTC requires that one parent be working at least 16 hours
each week, while eligibility for the childcare credit requires that no parent be
working less than 16 hours each week. It has been argued that it might be
desirable to offer the WTC and childcare credit to parents working less than
16 hours each week in order to encourage them to have at least some
involvement in formal employment, be it at very low hours. Hence, the fifth
option is:
5) Weakening the work requirement for the childcare credit from 16 hours
each week to any hours of employment
Such a modification could affect a considerable proportion of families. Some
12% of couples with children have one partner working at least 16 hours each
week and the other in formal employment at less than 16 hours each week,
while 6% of single parents are in paid employment but working less than 16
hours each week.
Changing the means-testing
Entitlement to WTC and the childcare credit will be means-tested on gross
family income. But limiting support for childcare costs only to mothers in
low-income families may not be desirable if the objective is to encourage all
mothers to work for reasons of economic efficiency or gender equity. Three
options are considered for extending eligibility for the childcare credit to
mothers in higher-income households:
6) Doubling the income threshold for means-testing the childcare credit from
£94.50 to £189 each week
7) Means-testing on the prime carer’s gross earnings for the childcare credit,
where the prime carer in a couple is the parent working the shortest hours
(typically the mother)9
8) Removing all means-testing for the childcare credit
9
In the estimation below, option 7 is implemented by removing the couple element from the
maximum credit entitlement so that the prime carer is effectively treated as a single person. In
addition, the 30-hour addition to the credit is only included if the prime carer is working at
least 30 hours each week.
91
Green Budget, January 2003
7.4 Potential numbers of eligible families
and budget costs, holding childcare and
employment constant
The impacts on the number of families eligible to benefit from the childcare
credit and the potential budget cost for each option described above are
presented in this section. It is assumed that employment and childcare
behaviour do not alter from current choices. The possible impacts of responses
in childcare and employment choices are discussed in the following sections.
The estimated numbers of eligible families and annual budget costs for the
baseline case of WTC and the options for modifications to the childcare credit
are presented in Table 7.3. It should be noted that the estimated numbers of
‘eligible families’ include not only those directly entitled to receive the
childcare credit, but all families that benefit indirectly because their
WFTC/WTC or child tax credit payment is higher on account of the fact that
they pay for childcare. The annual budget cost is the additional cost of the
WFTC/WTC and child tax credit programmes arising from the childcare
credit.
It should also be recognised that the results are presented in terms of the
number of families eligible to benefit rather than the actual number who
receive any benefit. In reality, the number actually receiving any subsidy may
be considerably smaller, as take-up of the WFTC is far from complete. Recent
estimates suggest that take-up rates are in the range of 62–65% among those
eligible, with considerably lower rates for couples with children (49–53%)
than for lone parents (77–83%).10 The estimated annual budget costs are based
on 100% take-up and will be smaller if not all eligible families claim the credit
to which they are entitled.
Just over 1.5 million families are estimated to be eligible for WFTC, with
115,000 benefiting from the childcare credit.11 Only a small proportion of
those eligible for WFTC are also eligible to benefit from the childcare credit
both because many couples have one parent (typically the mother) working
less than 16 hours and because many recipient families do not have any
10
Table 1 on page 4 of Inland Revenue, Working Families’ Tax Credit: Estimates of Take-Up
Rates in 2000–01, 2002. The take-up of the childcare tax credit has not been estimated.
11
The official statistics for the WFTC caseload as of August 2002 are 1,307,000 recipients of
WFTC and 170,200 recipients benefiting from the childcare tax credit (tables 1.1 and 1.3 of
Inland Revenue, Working Families’ Tax Credit Statistics: Quarterly Enquiry: United
Kingdom: August 2002, January 2003). The number of eligible families reported here may be
higher than the official number of recipients due to incomplete take-up, while the estimated
numbers eligible for the childcare credit may be lower than the official figure because of
under-reporting of childcare use in the Family Resources Survey data. Additional sources of
these discrepancies could be the differences in the time period considered or the sampling and
weighting procedures used in this analysis.
92
Childcare subsidies
eligible childcare costs. The average benefit from the childcare credit for
families receiving it is £30 each week,12 at an annual cost of £179 million.
Under WTC, 124,000 families will be eligible for the childcare credit, a
modest rise due to the adjustments listed above in the replacement of WFTC
with the WTC and child tax credit. The value of the average weekly care
subsidy for each family in receipt is estimated not to change from what it is
under WFTC, but the total budget cost of the care subsidy is estimated to rise
to £195 million.
Table 7.3. Number of eligible families and costs of the childcare credit:
childcare unchanged
Baselines:
WFTC
WTC
Options affecting eligible childcare
expenditure:
(1) Double ceilings for eligible childcare
expenditure
(2) Raise eligible proportion of childcare
costs to 100%
(3) Include informal childcare as eligible for
childcare credit
Voucher scheme for all types of care:
(4a) £1.80 per child per hour of care
(4b) £2.67 per child per hour of care
Number of
families
eligible to
benefit
(1,000s)
Average
weekly
benefit per
family
Annual
budget cost
115
124
£30
£30
£179m
£195m
125
£32
£205m
146
£42
£317m
188
£27
£263m
198
210
£30
£35
£305m
£382m
Option affecting work requirements:
(5) Reduce work requirement to any hours
135
£30
£208m
Options affecting means-testing for childcare
credit:
(6) Double income thresholds
169
£31
£270m
(7) Means test on prime carer’s earnings
425
£33
£739m
(8) Remove means-testing
584
£45
£1,370m
Notes: Families are eligible to benefit from the childcare credit if the WFTC/WTC or child tax
credit payment is higher on account of the fact that they pay for childcare. The average weekly
benefit is averaged over those entitled to benefit from childcare credit. The annual budget cost
is the additional cost of the WFTC/WTC and child tax credit programmes on account of the
childcare credit. All monetary values are in approximate April 2003 prices. The numbers of
families have been grossed to the national level.
Source: Author’s calculations using the Family Resources Survey (1998–99, 1999–2000 and
2000–01) and the IFS tax and benefit model, TAXBEN.
12
Official statistics indicate that families benefiting from the childcare tax credit receive an
additional £41 on average (table 8.1 of Inland Revenue, Working Families’ Tax Credit
Statistics: Quarterly Enquiry: United Kingdom: August 2002, January 2003). The lower
amount reported here may be due to incomplete take-up if families with a smaller childcare
credit claim are less likely to claim it.
93
Green Budget, January 2003
Options 1–4 show the impacts of modifying the amount of childcare
expenditure that is eligible for the childcare credit. Doubling the ceilings on
the amount of eligible childcare expenditure has little impact on the number of
eligible families and only a slightly larger impact on the annual cost. This is
because the vast majority of families spend less than the ceilings, although the
few who spend more could push the annual budget cost up by £10 million if
the ceilings were doubled. The second option – of raising the eligible
proportion of childcare costs from 70% to 100% – has a much greater impact
on the budget cost. It increases the number of eligible families to 146,000, as
some new families now pass the means test for the care credit, and raises the
estimated annual cost to £317 million, as all families currently eligible benefit
from a 43% rise in the childcare credit.13 The average childcare credit rises
from £30 per week for each family to £42 per week.
Making spending on informal care eligible for the credit (option 3) increases
the number of eligible families considerably, to 188,000, but has a more
modest impact on the annual budget cost, raising the annual tally to £263
million. Those using informal care spend less, on average, than those using
formal sources, and the average childcare credit for each family declines from
£30 each week to £27.
The most generous voucher scheme, considered in option 4b, is the most
expansionary of this group of options, raising the number of families eligible
for the childcare credit to 210,000 and the annual cost to £382 million. This is
a mixture of options 2 and 3. It is similar to option 2 because the hourly price
of £2.67 is, by construction, greater than 70% of the average informal costs
and allows a greater proportion of the costs to be covered. It is also similar to
option 3 in that the voucher can be used for both formal and informal care. It
is not surprising, therefore, that this voucher scheme raises the number of
eligible families by an amount slightly greater than option 3 and increases the
annual cost by an amount somewhat greater than option 2. The lower annual
budget cost (£305 million) and slightly lower number of eligible families
(198,000) under the less generous voucher scheme (option 4a) highlight how
limiting the voucher value can reduce the budget cost.
Reducing the work requirement from 16 hours each week to any hours (option
5) has moderate impacts. This may reflect the fact that families with a parent
in paid employment but working less than 16 hours each week are not likely to
be using childcare or spending large amounts on it.
The final three options in Table 7.3 (options 6–8) consider the effects of
modifying the means-testing for the childcare credit. Simply doubling the
thresholds for the means-testing has a moderate impact, increasing the number
of eligible families to 169,000 and raising the estimated annual cost to £270
million. On the other hand, moving to means-testing on the basis of the prime
carer’s earnings has substantial effects, raising the number of families eligible
for the childcare credit over three times to 425,000 and increasing the annual
budget cost by almost a factor of four, to £739 million. This is not surprising,
given that mothers in couples with both partners working at least 16 hours
each week are often considerably ‘poorer’ than their partners: on average, such
13
The rise is 30/70 (43%) corresponding to the movement from 70% to 100% of expenditure.
94
Childcare subsidies
mothers contribute only 38% to total family earnings, partly because 43%
work less than 30 hours each week. They are therefore likely to be eligible for
the credit as individuals if not as couples. The rise in annual cost is
proportionally greater than the rise in the number of eligible families because
families further up the income distribution tend to spend more on formal
childcare.14 Removing all means-testing for the childcare credit (option 8)
would increase the number of eligible families even further (to 584,000) and
would have a very large effect on the estimated annual cost. Indeed, the annual
cost would increase more than seven times from the baseline WTC case, to
well over £1 billion. Not only would the removal of all means-testing increase
the size of the eligible group, but the average value of the childcare credit
would rise considerably from the baseline of £30 per week per family to £45.
Overall, assuming no response in employment or childcare choices, the
options that alter the amount of eligible childcare expenditure, including
extending coverage to informal care, have moderate effects on the potential
number of families that benefit and on budget costs. Altering the meanstesting has more substantial effects, particularly any movement towards
means-testing on an individual rather than family basis. However, the next
section shows that if families respond to changes in policy by expanding their
childcare use, these conclusions need to be modified.
7.5 Allowing childcare responses (holding
employment constant)
Will childcare use and expenditure expand?
Modifying the childcare credit would create incentives for parents to alter
childcare choices even in the absence of any change in employment. In
particular, options that effectively reduce the price of childcare may lead to an
expansion in childcare use and expenditure. Parents may see greater use of
childcare as a way to work longer hours, to have more time free of their
children or to give their children the benefit of different types of care.
However, there are two important reasons why childcare use might not expand
in response to an increase in the generosity of the childcare credit:
•
Parents may feel that more time in paid-for childcare would be bad for
their children, even if it were effectively costless.
•
Sufficient formal childcare places or informal sources of childcare may
simply not be available to meet all the extra demand.
Yet even if the numbers of hours of childcare do not change, the amount spent
on care might rise:
14
Section 6.6.3 of G. Paull and J. Taylor with A. Duncan, Mothers’ Employment and
Childcare Use in Britain, Institute for Fiscal Studies, London, 2002.
95
Green Budget, January 2003
•
Parents may spend more money on childcare by choosing care options of
higher quality at a higher price.15
•
If the supply of formal childcare places is slow to respond to any increases
in demand, providers may raise their prices. The incidence of the care
subsidy is then said to fall in part on the providers because they gain part
of the benefit rather than the direct recipients.16 In the longer term and in
the absence of any barriers to new entry, new suppliers should enter the
market and bid down the price and parents should then receive the full
benefit of the subsidy.
•
The non-market nature of the provision of informal care generates large
incentives to expand the cost of this type of care without altering the hours
of care. In the informal arrangement, the amount of monetary transfers
between the parent and carer may be of no relevance, either because the
transfers can be returned in some other way (monetary or non-financial) or
because the parent and carer care sufficiently about each other that they
effectively operate as one household. Indeed, the provision of a large
proportion of informal care at no cost is evidence that financial payment
for care is not relevant for many parents and informal carers.17 Any
childcare subsidy for this type of care creates an incentive for both parent
and carer to report the greatest possible cost of the care (either through a
high hourly cost or through long hours) in order to maximise the amount
of the childcare credit that can be shared between them.
For these reasons, it is likely that any increase in the generosity of the
childcare credit would generate an expansion in the expenditure on childcare.
Whether there would also be an increase in the use and hours of care would
depend upon how much the demand for care and the supply of formal
childcare places and informal sources of care react to changes in the effective
price.
Scenarios for childcare expansion
How childcare use might respond to the modifications in the childcare credit is
not estimated here. Instead, the impacts of several scenarios for different
childcare reactions are explored. Many of these scenarios consider the upper
limits on the responses in childcare use and thereby provide approximate
upper bounds to the potential changes in the numbers eligible and budget
costs.
These scenarios for childcare expansions are applied to the four options that
would have the greatest potential impact on the use and cost of care: options
2–4b. Option 2 reduces the effective price of formal care from 30% of the cost
to zero. Option 3 includes informal care as eligible for the subsidy and
15
G. Paull and J. Taylor with A. Duncan, Mothers’ Employment and Childcare Use in Britain,
Institute for Fiscal Studies, London, 2002, find in chapter 7 that quality responses to variation
in the price of childcare may indeed be greater than quantity reactions.
16
A discussion of incidence is contained in Chapter 9.
17
An alternative explanation is that the carer actually enjoys caring for the child and does not
require financial compensation for doing so.
96
Childcare subsidies
generates the incentives described above to expand the cost and use of such
care. Options 4a and 4b contain incentives for both formal and informal care
to expand by reducing the effective price to zero for costs at or below £1.80
and £2.67 per hour for both types of care respectively.
It is initially assumed that families do not alter the type of care that they use,
but only increase the hourly cost or hours of care. This means that any
expansion is initially limited to those currently using formal care for option 2
and currently using formal or informal care for options 3, 4a and 4b. However,
it is obviously unrealistic to assume that parents would not alter the type of
care that they use in response to greater subsidies. It is particularly likely that
parents who do not currently use care may begin to do so. Therefore, three
progressively more expansionary scenarios are applied, first, only to current
users of childcare and, second, to all families fulfilling the work requirements
for the childcare credit:
a) The hourly cost rises to £2.67 (the current average for formal care)
b) The hourly cost rises to £2.67 and the hours of care per child rise to equal
the prime carer’s work hours
c) Total care expenditure rises to equal the ceiling on eligible expenditure
through any combination of increasing hourly price or weekly hours
As the expansion in scenario a is dependent upon some current usage of care,
it has identical impacts whether applied only to current users or to all working
families. Scenarios b and c are identical for the voucher schemes, as scenario b
implies the same maximum benefit as scenario c in these cases. Scenario c,
when applied to all families fulfilling the work requirement to be eligible for
the childcare credit, generates an upper limit on the costs of the childcare
credit assuming that no new families become eligible for the childcare credit
through changes in employment or family income.
It should be stressed again that these are upper limits and actual childcare use
and expenditure are very unlikely to approach these levels.
Potential eligibility and budget costs with an expansion in
childcare
The impacts on the number of eligible families and the annual budget cost
from each of these options and childcare expansion scenarios are presented in
Table 7.4. These estimates continue to assume that families do not alter their
employment choices in response to modifications in the childcare credit. It
should be noted, not least as a hint towards the conclusions, that the annual
budget costs are now presented in billions of pounds in contrast to the millions
of pounds in Table 7.3.
Option 2 increases the proportion of formal care costs that are covered by the
childcare credit from 70% to 100%. If those using formal care react to this by
expanding their hourly cost to £2.67, there would only be a moderate impact
on the number of eligible families and annual cost (scenario a). This is because
average hourly care costs are already quite high for this group. If, in addition,
current formal care users expand care hours to match the prime carer’s work
hours (scenario b), there would be a further moderate impact on the numbers
97
Green Budget, January 2003
eligible but a dramatic rise in the estimated annual budget cost, to £1.2 billion.
Allowing expansion by formal care users to reach the ceilings for eligible care
costs in the credit (scenario c) increases the number of eligible families to over
300,000 and would raise the estimated budget cost by over five times, to £1.8
billion.
Table 7.4. Number of eligible families and costs of the childcare credit:
with potential childcare responses
Number of families eligible
to benefit (1,000s)
Expansion scenario:
(a)
(b)
(c)
Annual budget cost
Expansion scenario:
(a)
(b)
(c)
(2) Raise eligible proportion of
No expansion: 146
No expansion: £0.3bn
childcare costs to 100%
formal users expand care
185
270
319
£0.5bn £1.2bn £1.8bn
all working families use care
185
1,849
2,163
£0.5bn £8.9bn £13.6bn
(3) Include informal childcare as
No expansion: 188
No expansion: £0.3bn
eligible for childcare credit
all current users expand care
607
742
871
£1.7bn £2.6bn £3.9bn
all working families use care
607
1,567
1,840
£1.7bn £5.4bn £8.2bn
(4a) Voucher scheme of £1.80
No expansion: 198
No expansion: £0.3bn
per child per hour of care
all current users expand care
526
746
746
£1.1bn £2.9bn £2.9bn
all working families use care
526
1,579
1,579
£1.1bn £5.9bn £5.9bn
(4b) Voucher scheme of £2.67
No expansion: 210
No expansion: £0.4bn
per child per hour of care
all current users expand care
609
898
898
£1.8bn £5.0bn £5.0bn
all working families use care
609
1,887
1,887
£1.8bn £10.3bn £10.3bn
Notes: Families are eligible to benefit from the childcare credit if the WFTC/WTC or child tax
credit payment is higher on account of the fact that they pay for childcare. The annual budget
cost is the additional cost of the WFTC/WTC and child tax credit programmes on account of
the childcare credit. All monetary values are in approximate April 2003 prices. The numbers
of families have been grossed to the national level. Expansion scenario a expands the hourly
cost to £2.67; scenario b expands the hourly cost to £2.67 and care hours to equal the prime
carer’s hours; and scenario c expands care expenditure to the ceiling on eligible expenditure.
Source: Author’s calculations using the Family Resources Survey (1998–99, 1999–2000 and
2000–01) and the IFS tax and benefit model, TAXBEN.
Any modification to the childcare credit that enticed those not currently using
care to begin to do so could have very large effects: more than half (56%) of
families fulfilling the work requirement report that they do not currently use
any childcare and 82% do not use any formal care. It is not surprising that the
scenarios where new working families may begin to use paid childcare in
response to the subsidy generate much larger impacts, raising the estimated
number of families eligible to benefit to around 2 million in the case of option
2. The annual budget costs are estimated to increase dramatically, to almost
£9 billion if hourly cost rises to £2.67 and hours of care rise to equal the prime
carer’s work hours (scenario b) and to over £13 billion if expenditure by all
rises to the maximum amount covered by the childcare credit (scenario c).
Returning to the cases where the care expansion is limited to current users, the
impacts are more dramatic if eligible care costs are extended to informal care
(option 3). Because initial hourly costs are so low (and a substantial proportion
free) for informal care, even increasing the hourly cost (scenario a) would
98
Childcare subsidies
increase the number of eligible families from 188,000 to 607,000 and the
estimated annual cost from under £0.3 billion to £1.7 billion. Because the
hours used also tend to be low, expanding the hours to match the prime carer’s
work hours (scenario b) would have a further substantial impact, particularly
on the annual budget cost, which is estimated to rise to well over £2½ billion.
Allowing expenditure to reach the ceilings for current users (scenario c) would
increase the number of eligible families to 871,000 and raise budget costs to
approach £4 billion.
Assuming all working families use childcare creates moderately lower
numbers of eligible families and annual budget costs under option 3 than
option 2, because option 3 provides a maximum credit that is only 70% of that
in option 2. The increase in potential costs by assuming that new families use
paid childcare over expansion only by current users is of a much smaller
magnitude when eligible childcare costs are limited in this way.
The less generous voucher scheme (option 4a) has very similar effects to the
WTC extended to include informal care (option 3) when expansions are
limited to an hourly cost of £2.67 and hours equal to the prime carer’s work
hours (scenario b). Yet the advantage of the voucher scheme is that there
would be no further impact on the annual budget cost if childcare expanded
beyond this point, because childcare expenditure under this scenario has
already exceeded the maximum eligible amounts for the voucher scheme. If
current users expanded their use further to the ceiling on eligible expenditure
(scenario c), the weekly ceilings from the current system used in option 3
mean that option 3 is estimated to cost £1 billion more than the voucher
scheme (option 4a) (£3.9 billion compared with £2.9 billion). If all working
families used care to the ceiling on eligible expenditure, option 3 is estimated
to cost over £2 billion more than the voucher scheme (£8.2 billion compared
with £5.9 billion). However, the more generous voucher scheme (option 4b)
has implicit ceilings on expenditure that are, on average, far higher than those
in the current system. Consequently, the annual budget costs are considerably
greater for this voucher scheme than for option 3 under both expansion
scenarios b and c.
This emphasises the point that if childcare subsidies contain incentives for
substantial expansions in childcare use, the ceilings on the subsidy may play
an important role in determining the budget cost. Indeed, given the extremely
large magnitude of the estimated costs, political acceptability for consideration
of some of these options is likely to require more stringent ceilings on eligible
care expenditure to ensure that potential costs would be contained to
reasonable levels.
Yet it should also be stressed that these estimated impacts of childcare
responses are upper bounds on potential claimant numbers and costs. In the
short term, lack of availability is likely to constrain the effects to much more
limited rises, while less than complete take-up of the credit and consideration
of the desirability of greater use of childcare for the children may mean that
the costs do not reach these levels even in the longer run. The estimates in
Tables 7.3 and 7.4 are also premised on the assumption that families do not
alter their employment behaviour. Relaxation of this assumption is discussed
in the following section.
99
Green Budget, January 2003
7.6 Considering employment responses
With the exception of option 5,18 the modifications to the childcare credit
affect employment choices indirectly by reducing the effective cost of
childcare for working families. If childcare is required to facilitate
employment, a reduction in the effective hourly price of care is analogous to
an increase in the hourly wage rate for the prime carer. It therefore has the
same two opposing effects on labour supply:
•
A substitution effect: if the effective hourly price of childcare falls, the
financial return to each hour of work rises and parents may choose to work
more.
•
An income effect: a fall in the cost of childcare makes any current use of
care less expensive, raising the effective income of parents, which may
lead them to feel that they can afford to reduce their hours of work.
Which of these effects is likely to dominate the overall labour supply response
cannot be predicted without additional empirical estimation. Nevertheless,
several points can be noted:
•
The childcare credit only applies to parents working more than 16 hours
each week. Any increase in its generosity cannot entice parents to leave
employment or reduce hours below 16 each week through the income
effect because parents would then not benefit from the subsidy. Hence, an
increase in generosity in the childcare credit can potentially raise, but not
reduce, the proportion of parents who are working 16 or more hours each
week. Families enticed to enter employment must do so at hours and
earnings levels that make them eligible for the childcare credit.
•
The positive substitution effect on work hours can only operate if
additional childcare is both required and available to facilitate an
expansion in work hours. For example, parents of older school children
may not need additional childcare to work longer and therefore would not
react to a change in the effective childcare price.
•
The positive substitution effect on work hours will not operate for families
already receiving the maximum level of care subsidy. However, if the
maximum value of the subsidy is directly linked to the work hours of the
prime carer (as in the voucher schemes described above), there will be a
positive substitution incentive for all families.
•
Given the relative lack of responsiveness in men’s working hours to
changes in the wage rate,19 any employment response is more likely to be
18
Option 5 reduces the work requirements for the WTC and childcare credit from 16 hours
each week to any hours of employment. This encourages parents either to reduce their work
hours to below 16 each week or to enter employment in order to work the shorter hours.
Hence, this modification has incentives both to reduce weekly hours for those already working
and to increase the proportion of parents in employment.
19
For example, see R. Blundell and T. MaCurdy, ‘Labor supply: a review of alternative
approaches’, chapter 27 of Handbook of Labor Economics, Elsevier Science, 1999.
100
Childcare subsidies
seen in changes in the work choices of the mother than in alterations in the
father’s hours of work.
•
Any reduction in hours of employment due to the income effect could lead
to a decline in the number of childcare hours.
If employment does increase (or fall), it is likely, although not automatic, that
childcare use would also move in the same direction in line with the change in
working hours. In terms of the maximum potential budget costs (where all
working families purchase care to the credit ceilings) presented in Table 7.4,
employment responses may raise these costs to the degree that families enter
employment or increase their weekly working hours to 16 or more. Given that
approximately half of all families do not currently fulfil the work requirements
for the childcare credit, the impacts could be large if a substantial proportion
of these families increased their labour supply in response to a more generous
childcare credit. But past empirical evidence suggests that labour supply
responses are likely to be much more modest. For example, simulations of the
impact of the introduction of the WFTC estimated that the employment
participation rate for single mothers would rise by only 2 to 3 percentage
points.20 Yet even if subsidy costs did increase substantially as a result of
increased work hours or employment participation, this would be directly
paying for achieving the objective of improving the labour market
involvement of parents.
7.7 Impacts across families
As highlighted in the initial discussion of the employment choices of families
and the use of childcare by working families (Section 7.2), expansions in the
childcare subsidy could have diverse impacts for different types of families. In
this section, the main distributional consequences of the childcare subsidy
modifications are briefly summarised.21
The childcare credit in the WTC will benefit different types of families to
varying degrees:
•
Single parents are much more likely to be eligible to benefit from the
childcare credit than couples with children: single parents are estimated to
constitute some 67% of the families who are eligible for the childcare
credit.
•
Families with a pre-school child are more likely to be eligible to benefit
from the childcare credit than those with only school children: over half of
the eligible families are estimated to have at least one pre-school child.
•
Families with more than one child are slightly less likely to be eligible to
benefit than those with a single child.
20
R. Blundell, A. Duncan, J. McCrae and C. Meghir, ‘The labour market impact of the
working families’ tax credit’, Fiscal Studies, 2000, vol. 21, pp. 75–103, and A. Duncan, G.
Paull and J. Taylor, ‘Mother’s employment and the use of childcare in the UK’, Institute for
Fiscal Studies, Working Paper no. 01/23, 2001 (www.ifs.org.uk/workingpapers/wp0123.pdf).
21
The tables and figures for this summary are available on request from the author.
101
Green Budget, January 2003
•
Most families who are eligible to benefit from the childcare credit have
income levels in the middle of the income distribution.22
The options for expansion would distribute benefits unevenly across family
types. In terms of the percentage change in each type who would become
eligible to benefit from the childcare credit:
•
Expanding the proportion of formal care that is covered by the subsidy
(option 2) would generate moderately greater benefits for couples over
single parents, for those with pre-school children and for those above
median incomes.
•
Expanding the credit to include informal care (options 3, 4a and 4b) would
generate greater benefits for couples, families with only school children
and families with more than one child if the use of care expands.
•
Expanding the credit to include informal care (options 3, 4a and 4b) would
exaggerate the greater benefits drawn by those in the middle of the income
distribution, particularly if the use of care expands.
•
Those at the very bottom of the distribution would be most likely to
benefit if working families not currently using care began to do so if the
credit were expanded to include informal care (options 3, 4a and 4b).
•
Those in the top of the income distribution would benefit the most if
working families not currently using care began to do so if either (i) the
proportion of formal care covered by the subsidy were expanded to 100%
(option 2) or (ii) the more generous voucher scheme were introduced
(option 4b).
•
Modifications to the means-testing (options 6 to 8) would generate greater
benefits for couples.
•
Means-testing on the prime carer’s earnings (option 7) would have the
greatest benefits for those around the eighth decile in the income
distribution, but complete removal of means-testing (option 8) would
generate substantial benefits to those at the very top of the income
distribution.
7.8 Conclusions
Government support for childcare costs for working families has increased
considerably over the past decade, and recent initiatives continue to move
gradually in the direction of more generous subsidies. Yet the more radical
possibilities for changes in the scope of support, such as covering informal
care or increasing support further up the income distribution, remain on the
fringes of policy consideration. Concern for a sudden explosion in the budget
cost of the more sweeping reforms may be one reason for hesitation. In
22
This is due, in part, to the fact that eligibility for the childcare credit requires all parents to
be in employment for a minimum of 16 hours each week, generating a reasonably high level
of family earnings. It is also due to the fact that the likelihood of using formal childcare is
greater for families with higher incomes.
102
Childcare subsidies
particular, some proposals contain incentives for an unknown and potentially
large reaction in the childcare choices of parents that could prove very
expensive to subsidise.
This chapter has considered the potential costs of several options for
increasing the generosity of the childcare credit. It has included an analysis of
some of the most expensive scenarios for an expansion in childcare use and
expenditure. It is likely that the corresponding cost estimates overstate the
likely outcomes because:
•
it is very unlikely that any reforms to the childcare credit would invoke
such extreme changes in behaviour, especially in light of reports of a
shortage of childcare places;
•
the cost estimates assume complete take-up of the credit, while actual takeup rates are currently much lower.
Yet consideration of such upper bounds has shown that whether the potential
costs fall within the realms of an affordable policy depends to a large degree
on how parents react in their childcare choices:
•
Expanding the coverage of the childcare credit to a greater proportion of
formal care costs or to include informal care does not have a substantial
impact on budget costs if childcare choices remain unchanged.
•
But if childcare choices responded to extreme levels, the budget costs of
the childcare credit could escalate enormously to well beyond what is
likely to be deemed affordable.
•
Modifications to the means-testing for the childcare credit are less likely to
alter childcare behaviour substantially and are likely to impose large, but
perhaps feasible, increases in budget cost.
In addition, the desired objectives of enhancing childcare use and employment
rates among parents may not accompany a rise in budget costs:
•
There are good reasons to believe that expenditure on both formal and
informal care could rise in response to increases in the childcare credit’s
generosity without any substantial change in childcare use or hours.
•
The employment incentives of increasing the level of the care subsidy are
mixed: some parents could be encouraged to enter employment or extend
their hours, while others might react by reducing hours.
Therefore, consideration of any expansion in the childcare credit may require
modifications in the current design in order to ensure that the costs cannot
spiral beyond acceptable levels and to enhance the effectiveness of the subsidy
in achieving childcare and employment objectives. Such modifications could
include:
•
lowering the current weekly ceilings on childcare expenditure;
•
relating the value of the credit to childcare hours (as in an hourly voucher)
to ensure that the more substantial potential rises in costs are matched by
increases in the use and hours of care;
103
Green Budget, January 2003
•
relating the maximum number of eligible care hours to the prime carer’s
hours of employment in order both to target resources to those with
greatest need in terms of facilitating employment and to improve the
incentives to work longer hours;
•
relating eligibility for the credit to the employment and earnings of the
prime carer (typically the mother) in order to maximise the positive
employment incentives for the parent most likely to be responsive.
As it currently stands, the childcare credit does not cost much simply because
it does not help many families. Providing greater support for childcare costs
for working parents to such an extent that it might encourage more parents
into paid employment would inevitably increase the size of the financial
burden of the subsidy. If such an objective is desired, a careful redesign of the
childcare credit and consideration of the reactions in childcare use and
expenditure may be required to ensure that the costs are contained to an
affordable level that would make greater support politically feasible.
Gillian Paull
104
8. Measuring public sector
efficiency
The government has staked its political credibility on delivering significant
and noticeable improvements to public services. Over the period from April
1999 to March 2006, the resources allocated to public services are set to
increase substantially: NHS spending is forecast to increase by 7.3% per year
in real terms, spending on transport by 6.3% and spending on education by
5.9%.1 The extent to which these increased resources are translated into
improved outcomes for users will depend on the efficiency, or productivity, of
public service providers. In addition, the extent to which public sector
productivity improvements materialise will have implications for the
government’s aim of increasing the rate of productivity growth of the UK
economy.
The 2002 Spending Review set out around 130 Public Service Agreement
(PSA) performance targets covering areas such as health, education and crime.
The PSA targets cover both outcomes, such as NHS waiting times, and the
efficiency with which those outcomes are delivered. For example, Department
of Health objective 12 states that ‘value for money in the NHS and personal
social services will improve by at least 2% per annum, with annual
improvements of 1% in both cost efficiency and service effectiveness’ and
Home Office objective 10 aims to ‘ensure annual efficiency gains by the
police of at least 2%’.2
These targets raise issues about how the efficiency of individual service
providers is measured, and also about whether the set of efficiency measures
produced can be used to give service providers incentives for improved
performance. Reliable measures of both outcomes and provider efficiency are
particularly important if they are to be used to target resources to more
efficient providers, to act as a motivational tool or to detect failing providers.
In April 2000, HM Treasury’s Public Services Productivity Panel published a
report detailing a new approach to measuring the efficiency of the police.3 The
wider aim was to use the efficiency measures as part of an incentive system to
1
HM Treasury, 2002 Spending Review, Cm. 5570, London, 2002 (www.hmtreasury.gov.uk/Spending_Review/spend_sr02/spend_sr02_index.cfm?); HM Treasury, Public
Expenditure Statistical Analyses 2002–03, Cm. 5401, London, 2002 (www.hmtreasury.gov.uk/mediastore/otherfiles/pesa_2002to2003.pdf); authors’ calculations. For
comparable historical figures, see table 3.4 in A. Dilnot, C. Emmerson and H. Simpson (eds),
The IFS Green Budget: January 2002, Commentary no. 87, IFS, London, 2002
(www.ifs.org.uk/gbfiles/gb2002.shtml).
2
‘Public Spending and Services: Links to Departmental Performance Documents’, www.hmtreasury.gov.uk/Documents/Public_Spending_and_Services/publicservice_performance/pss_p
erf_table.cfm.
3
C. Spottiswoode, Improving Police Performance: A New Approach to Measuring Police
Efficiency, HM Treasury Public Services Productivity Panel, London, 2000 (www.hmtreasury.gov.uk/mediastore/otherfiles/231.pdf).
105
Green Budget, January 2003
improve police performance. Police authorities were to be placed in efficiency
bands, which would be used to determine future performance targets and
inspection and review arrangements. It was also suggested that the proposed
efficiency measurement techniques – namely, ‘stochastic frontier analysis’ and
‘data envelopment analysis’ – might have wider applicability within the public
sector.4 The Office of the Deputy Prime Minister has also commissioned
research into the use of these techniques in measuring the cost-effectiveness of
local authorities.5 This chapter outlines the techniques that were proposed and
assesses their suitability for measuring the efficiency of public sector
organisations.
8.1 Efficiency measurement in the public
sector
Measuring an organisation’s efficiency is about the relationship between the
outputs it produces and the inputs it uses. An efficient organisation would be
one that produces the maximum possible outputs given its inputs, or one that
produces a certain level of output with the minimum amount of inputs. The
process of trying to measure an organisation’s efficiency can therefore be
broken down into three steps. First, its inputs and outputs need to be defined
and measured. Secondly, it is necessary to define what is feasible – in other
words, what outputs could be achieved for any given set of inputs. Finally, the
organisation’s actual inputs and outputs are compared with the set of feasible
inputs and outputs. At this stage, one of two questions can be asked: ‘Is it
feasible to achieve superior outputs, given the set of inputs being used?’ or ‘Is
it feasible to use less inputs to achieve the same outputs?’. The way the first
two steps are carried out will typically be highly influential on the outcome of
the third.
Definition and measurement of outputs, inputs and
environmental factors
Efficiency measurement is relatively straightforward for an organisation
producing one type of output with one type of input. But most organisations –
public and private – produce a wide range of outputs and use numerous inputs.
In the case of a private firm selling its output in a competitive market,
different outputs can be aggregated by using the observed prices.6 But public
sector organisations usually produce goods that are provided either free at the
point of use or at a price that is not determined by market forces. This makes it
4
HM Treasury, ‘Improving Police Performance: A New Approach to Measuring Police
Efficiency’, Press Release PSP4, 17 April 2000.
5
Office of the Deputy Prime Minister, Local and Regional Government Research Programme
Newsletter 2002–03 (www.local.dtlr.gov.uk/research/02.htm).
6
Under certain assumptions, prices will reflect market or buyers’ valuations of the outputs.
They therefore act as a natural set of weights that can be used to aggregate outputs into a
single measure. See, for example, W. E. Diewert, ‘Fisher ideal output, input and productivity
indexes revisited’, Journal of Productivity Analysis, 1992, vol. 3, no. 3, pp. 211–48.
106
Measuring public sector efficiency
very difficult to define the aggregate output of a public service provider such
as a school, hospital or police force. Inputs, such as hospital beds, are usually
easier to deal with, as prices are more often observed.
Another difficulty arises because some inputs are not under the control of an
organisation. These can include environmental variables such as the
characteristics of individuals using the service – for example, their underlying
health – that may have an impact on measured efficiency. It is therefore
necessary to find a way to take account of their effects when comparing
organisations.
Defining what is feasible
With an appropriate set of inputs and outputs at hand, the next task is to define
the efficient set of inputs and outputs against which an organisation can be
compared. This is, however, unknown. The procedure therefore is to compare
an organisation with an ideal comparator constructed from information on
other organisations operating in the same field (and with similar size and
environmental factors etc.).7 In practice, this is often difficult, as similar
organisations may be few and far between, especially in the public sector.
Environmental differences are particularly difficult to control for in this
respect. For example, schools will differ in the average ability of children in
their catchment area and ambulance service response times will vary
according to population density. If these environmental factors are taken into
account when selecting the comparison group, then the group will often be
rather small or even empty. If they are not, then any efficiency measurements
that emerge may simply reflect differences in these environmental factors.
Measuring efficiency
Setting the difficulties of finding comparable organisations aside, two
particular technical methods were put forward in the Public Services
Productivity Panel report as useful ways of deriving efficiency measures for a
group of ‘similar’ organisations.
The two methods that have been proposed are stochastic frontier analysis
(SFA) and data envelopment analysis (DEA). It is important to ascertain
whether these techniques are really robust and general enough to achieve the
purposes for which it has been suggested they are used.8 To look at this, we
begin by describing the two techniques. Consider Figure 8.1. The dots
represent observed input and output combinations for six organisations. The
two sets of lines (the solid curved line and the dotted line made up of lots of
straight segments) are ‘frontiers’ and show the maximum output that could be
7
Of course, given the information available, this comparator may in itself be inefficient
compared with what is possible.
8
For a discussion of these issues with particular reference to the Spottiswoode Report, see M.
Stone, ‘How not to measure the efficiency of public services (and how one might)’, Journal of
the Royal Statistical Society A, 2002, vol. 165, issue 3, pp. 405–34, and the discussion in the
same volume.
107
Green Budget, January 2003
produced for each level of input. The two frontiers represent the outcomes of
the two methods of measuring efficiency, SFA and DEA.
Figure 8.1. SFA and DEA
Output
SFA frontier
DEA frontier
E
} noise
{
B
noise
inefficiency
{
D
F
C
A
Input
Stochastic frontier analysis
Stochastic frontier analysis uses statistical methods to fit a frontier like the
solid curve in Figure 8.1. The idea is to identify the relationship between
output and input(s) whilst allowing for two types of deviation from this
relationship. One is statistical ‘noise’ – in other words, random variations in
the data caused by inaccuracy in the measurement of output and by other
errors. This first type of deviation is assumed to be zero on average, so that, on
average, output is measured accurately.9 The second type of deviation is a
measure of inefficiency. It is one-sided: if a firm were fully efficient, it would
be zero, and the more inefficient the organisation is, the more negative the
deviation. These two types of deviation from the efficient frontier are shown
in the figure by the curly brackets for organisations D and E. In this case,
organisations B and E are classed as efficient as they lie above the frontier and
organisations A, C, D and F are inefficient to some degree. The extent to
which an organisation’s total deviation from the frontier is designated to be
noise versus inefficiency depends on the choices made about the joint
distributions of the two components.
Data envelopment analysis
Data envelopment analysis is a non-statistical approach to the problem of
efficiency measurement. Put simply, it takes data on organisations’ outputs
and inputs, and measures the efficiency of a particular organisation by its
distance from the ‘outer envelope’ of the data. This outer envelope is shown in
Figure 8.1 by the dashed line for the case where there are assumed to be
9
As long as this noise is not correlated with either the inputs or the second type of deviation, it
is of no particular interest.
108
Measuring public sector efficiency
variable returns to scale. With this technique, all deviations from the efficient
frontier are classed entirely as inefficiency. In the figure, the solid arrow
represents a measure of organisation D’s inefficiency. Organisations A, B and
E are measured as efficient and organisations C, D, and F as inefficient. It is
worth noting that this procedure (and this variable-returns-to-scale version of
it in particular) can designate an organisation as completely efficient simply
because it produces more of a particular output than other organisations. In
this single-output example, the organisation that produces the most will find
itself on the efficient frontier simply because there is no larger organisation
with which to compare it.
Measurement error and the form of the frontier
Stochastic frontier analysis requires a great deal of knowledge, both about the
shape of the frontier and the distributions of the two types of deviation, if it is
to yield a useful answer. The choice that is made about the shape of the
frontier and the distributions of the deviation components can have significant
effects on the efficiency rankings and absolute efficiency measures generated.
These choices can be pretty much arbitrary: economic theory often provides
little information about the shape of the frontier, and the data can be
uninformative about the distributions of the two types of deviation. Failing to
measure the inputs accurately can further complicate the task of correctly
identifying organisations’ relative efficiency.10
Data envelopment analysis does not require any assumptions about the shape
of the frontier or about statistical distributions. But as the whole approach, by
definition, focuses on extreme observations, it is very sensitive to
mismeasurement. For example, in Figure 8.1, if the output of organisation E
were inaccurately recorded and overstated, inclusion of E in the frontier would
mean that the frontier was mismeasured and that the inefficiency of
organisations such as F would be overstated. Furthermore, actual organisations
are generally compared with hypothetical organisations (for example, the point
on the DEA frontier with which organisation D is compared is a hypothetical
organisation), which can mean that results are fragile if there are few real
observations from which to determine the potential performance of these
hypothetical organisations.
How many inputs and outputs to include
Both approaches will give results that depend upon the choice of inputs and
outputs considered. An important issue is how many to include. As the number
of inputs and outputs measured increases, the task of measuring efficiency
becomes rapidly more difficult.
In the SFA approach, one problem is trying to choose appropriate forms for
the frontier where the data are very sparse and are barely informative about the
appropriate choice. Another is specifying how the many noise and inefficiency
components are distributed.
10
More generally, the problem is one of measurement error in the explanatory variables. See
P. W. Bauer, ‘Recent developments in the econometric estimation of frontiers’, Journal of
Econometrics, 1990, vol. 46, pp. 39–56.
109
Green Budget, January 2003
In the DEA context, as the number of outputs increases, it turns out that the
measured efficiency of the organisations cannot go down and, in the extreme,
all organisations will end up being regarded as 100% efficient.11 The intuition
for this is quite straightforward: if you specify outputs very finely, you will
eventually end up defining one output unique to each organisation, which –
not surprisingly – it is uniquely efficient at producing. Choosing the level of
aggregation of inputs and outputs is therefore an essential aspect of the
exercise. There is no really good way of knowing how far or how little to
disaggregate, yet the results will be heavily influenced by this choice. In order
for these methods to yield useful results, the number of dimensions must be
kept fairly moderate, and rules of thumb abound – for example, ‘the number of
organisations must be more than three times the combined number of inputs
and outputs’. Decisions must therefore be made about which inputs and
outputs to ignore, which to include and which to aggregate.
SFA and DEA in practice
Applications of these techniques have illustrated the sensitivity of results to
the methods used to measure efficiency. A relatively recent application of
statistical methods in measuring public sector efficiency can be seen in The
World Health Report 2000 – Health Systems: Improving Performance,12
which compares the efficiency of different countries’ health systems. A
scientific peer-group report on the methods employed in the paper was
commissioned.13 The report raised a number of the general issues discussed
above, as well as a large number of reservations specific to the World Health
Organisation’s application of the methods. In particular, the group questioned
the publication of league tables (page 123) based on the methods and data
used. Another application is work comparing the efficiency of hospital trusts,14
in which the authors highlight the sensitivity of the results to the efficiency
measurement techniques used.
8.2 Conclusions
The obvious practical question arises of how concentrating on a subset of
performance or input indicators might affect organisations’ incentives.
Clearly, formulaic application of these methods could have implications for
providers’ resource allocation, leading to the underprovision of outputs that
are not included in the efficiency measurement exercise. Similarly, if certain
organisations find it relatively expensive to produce certain outputs, they may
11
T. Nunamaker, ‘Using data envelopment analysis to measure the efficiency of non-profit
organizations: a critical evaluation – reply’, Managerial and Decision Economics, 1985, vol.
6, pp. 50–8.
12
World Health Organisation, The World Health Report 2000 – Health Systems: Improving
Performance, Geneva, 2000 (www.who.int/health-systems-performance/whr2000.htm).
13
‘World Health Report 2000 Consultation and Peer Review’, www.who.int/health-systemsperformance/consultation.htm.
14
D. Dawson, R. Jacobs and A. Street, ‘Comparing the Efficiency of NHS Hospital Trusts’,
www.niesr.ac.uk/event/jacobs.pdf.
110
Measuring public sector efficiency
tend to redirect resources to other outputs. This has obvious implications for
universal service obligations.15 More generally, trade-offs exist between
having large numbers of targets or a few, in terms of accurate performance
measurement, the costs of gathering and analysing performance information,
the incentives given to public service providers, and transparency and
accountability in public service provision.
The outcomes of DEA and SFA exercises, in the form of rankings or league
tables of organisations or other efficiency scores, are sometimes used in a
second-stage analysis that tries to uncover the reasons for these measured
differences in performance. Typically, the DEA or SFA efficiency scores are
regressed on factors that may affect performance – for example,
environmental factors that are thought to be under the organisations’ control.
Of course, this presupposes that the efficiency scores are correct and, for the
reasons discussed above and others specific to any particular application, they
may not be. Ultimately, very many auxiliary technical judgements are required
when implementing either SFA or DEA techniques. While the methods are
potentially useful, the results need to be treated with caution, as should their
application as a motivational tool.
Ian Crawford, Alexander Klemm and Helen Simpson
15
See A. Chesher, ‘Discussion on the paper by Stone’, Journal of the Royal Statistical Society
A, 2002, vol. 165, issue 3, pp. 423–4.
111
9. The distributional effects of fiscal
reforms since 1997
The government has carried out many reforms to the tax and benefit system
since coming to power in May 1997. We estimate that the overall effect of the
changes enacted or announced to date will be an annual cost to the exchequer
of around £1.6 billion in 2003–04. This is the difference between a large set of
revenue-raising reforms and a slightly larger set of costly changes. These
reforms will not have affected all groups in the population equally. Rather, the
gains or losses of different people will depend on individual characteristics
such as income level, age and family circumstances.
The aim of this chapter is to gain a deeper understanding of how tax and
benefit reforms have affected different groups in the population. The
discussion will also cover the issue of how the burden of ‘taxes on businesses’
has changed. But, in considering this, it is important to remember that
individuals ultimately pay all taxes. So our main focus is how the reforms
have affected people at different points in the income distribution.
After a first section that lists the main reforms made since 1997, Section 9.2
begins the detailed distributional analyses. There, we mainly consider ‘direct’
taxes on labour income and unearned income and ‘indirect’ taxes on
expenditure – the payment of which can be allocated relatively
straightforwardly to households according to their earnings and their
expenditures – as well as receipt of benefits, credits and tax credits.1 We then
discuss, in Section 9.3, some of the limitations of this analysis. These include
both conceptual limitations (in terms of how we allocate tax payments to
particular groups in the population) and data limitations (which constrain the
range of taxes we can consider). Section 9.4 looks at ‘taxes on business’, while
Section 9.5 considers the distributional consequences of changes to stamp duty
on domestic property. Section 9.6 concludes.
9.1 What fiscal reforms?
A helpful benchmark when considering the distributional effects of the fiscal
reforms made since Labour came to power is their overall impact on the
government’s finances. Table 9.1 reports an estimate of the effect on revenues
in 2003–04 of the reforms, compared with the revenues that would have been
expected if the tax system implied by the November 1996 Budget had been
retained (with benefit rates and expenditure taxes uprated in line with inflation
1
This section updates previous IFS work: M. Myck, Fiscal Reforms Since 1997, Briefing
Note no. 14, IFS, London, 2000 (www.ifs.org.uk/taxben/fiscalreform.pdf), and T. Clark, M.
Myck and Z. Smith, ‘Fiscal reforms affecting households, 1997–2001’, chapter 5 of T. Clark
and A. Dilnot (eds), Election Briefing 2001 IFS, London, 2001
(www.ifs.org.uk/election/ebn5.pdf).
112
Distributional effects of fiscal reforms since 1997
Table 9.1. Revenue effects in 2003–04 of changes to taxes and benefits
made since 1997
Net revenue
raised for
exchequer
Total income tax
Of which:
Married couple’s allowance
Income tax rates and personal allowances
Dividend tax credits
Net revenue
cost to
exchequer
£0.3bn
£3.3bn
£10.4bn
£6.5bn
Total National Insurance
Of which:
Employee contributions
Employer contributions
Self-employed contributions
£5.2bn
Total indirect taxes
Of which:
VAT
Tobacco taxation
Road fuel duties
Alcohol taxation
Insurance premium tax
Vehicle excise duty
£8.3bn
Total stamp duties
Of which:
Changes to rates for properties
£2.1bn
£2.0bn
£1.5bn
£0.6bn
£0.0bn
£2.6bn
£4.9bn
£0.2bn
£1.2bn
£1.1bn
£2.0bn
Total corporation tax
Of which:
Changes to rate structure
Other corporation tax changes
£2.8bn
£4.4bn
£1.6bn
Total change in cost of benefits / credits / tax credits
Of which:
Mortgage interest relief at source
Working families’ tax credit (TC) + Disabled
persons’ TC + Working TC + Child TC
Child benefit (and non-attributable child-based
reforms)
Pensioners’ package (winter allowance, basic
pension and minimum income guarantee increases)
Overall total
£14.3bn
£3.4bn
£8.6bn
£1.7bn
£6.1bn
£ 1.6bn
Notes: All costings have been deflated to 2003–04 using nominal GDP growth published by the Office
for National Statistics and (for projections) HM Treasury. The totals include all measures, not just the
taxes and benefits costed in detail in the table. Some taxes and benefits have been reformed more than
once since 1996–97, which means that they may score as both exchequer gains and exchequer losses.
The figures in the detailed breakdowns in this table are approximate. In some instances, it is not possible
to break down the cost of measures introduced into the categories given in the table. For example,
increases to child premiums in both income support and WFTC are often grouped together in costings
published by the Treasury. In such cases, the effect of the changes is either attributed to the category
deemed likely to be responsible for the greater part of the cost, or added to child benefit (and nonattributable child-based reforms). The difficulty with separating the effects of some measures also
explains why the WFTC, DPTC, WTC and child TC are considered as one category.
Sources: HM Treasury, Financial Statement and Budget Report and Pre-Budget Report, various years.
113
Green Budget, January 2003
and the parameters of the income tax system uprated in line with statutory
rules). The table also lists the costs or revenues raised by certain specific
measures that had large effects on the government budget. The measures
considered include some that were announced by the Conservative
government before May 1997 and which Labour chose to implement, such as
the tobacco and fuel duty escalators, which were originally announced in the
Budgets of 1993.
The table shows that the net effect of all policy changes in the tax and benefit
systems will have been a small fiscal loosening of around £1.6 billion
compared with what the government’s budgetary position would have been
had the system implied by the November 1996 Budget been retained. This
figure is actually the difference between a large set of revenue-raising
measures (around £51.7 billion) and a slightly larger set of costly reforms
(around £53.3 billion). For example, Table 9.1 shows that costly changes to
income tax, such as the introduction of the 10p starting rate and reduction of
the basic rate, have been almost entirely offset by revenue-raising changes to
the same tax. In contrast, by next year, the exchequer will have gained
significant net revenues from changes to National Insurance2 and indirect
taxes (such as road fuel duties), plus a smaller amount from stamp duties. The
tax changes overall raise around £12.7 billion, in effect paying most of the
£14.3 billion projected net cost of benefit increases and the creation of new
credits and tax credits.
The reforms listed in Table 9.1 only include fiscal changes that involve direct
financial transfers to or from the state. In the second half of its first term in
office, the Labour government also began to devote significant funds to extra
spending on the National Health Service, and on education and other public
services. Much of this spending was financed from extra tax revenues that
have not been due to the discretionary policy changes considered in Table 9.1.
Taxes as a share of national income in 2003–04 are forecast by the Treasury to
be 39.3%, compared with 37.4% in 1996–97.3 This increase of 1.9 percentage
points is equivalent to £20.8 billion in 2003–04 terms. Analysis of the increase
to 2001–02 shows that much of it is increased income tax receipts as a share of
national income. Likely causes of the overall increase include changes in the
composition of national income, and the level of the oil price.4
In this chapter, we will not attempt to consider the distributional effects of
spending on public services, nor the impact of changes in tax take or benefit
expenditure that are not due to policy changes: we focus our attention on the
effects of discretionary changes in financial transfers. In the next section, we
consider the distributional effects of those reforms where the financial transfer
can most straightforwardly be allocated to particular households. The latter
2
For details of the changes to National Insurance, see Chapter 5.
3
HM Treasury, Pre-Budget Report 2002, Cm. 5664, London, 2002 (www.hmtreasury.gov.uk/pre_budget_report/prebud_pbr02/prebud_pbr02_index.cfm).
4
For a fuller discussion, see chapter 2 of A. Dilnot, C. Emmerson and H. Simpson (eds), The
IFS Green Budget: January 2001, Commentary no. 83, IFS, London, 2001, and chapter 3 of
T. Clark and A. Dilnot (eds), IFS Election Commentary, Commentary no. 84, IFS, London,
2001.
114
Distributional effects of fiscal reforms since 1997
part of the chapter discusses some of the reasons why it is more difficult to
model the distributional effects of changes in ‘business taxes’ and stamp
duties.
9.2 The distributional impact of reforms
directly affecting households
In the previous section, we listed the main reforms to taxes and benefits that
have occurred since 1997. In this section, we use TAXBEN, the IFS
simulation model of the tax and benefit system, to assess the distributional
impact that followed from some of them.
The taxes and benefits that we include in the analysis are those levied directly
on incomes or on personal expenditures. We can allocate the payment of these
taxes and benefits to households using data on the incomes and expenditures
of household members. The direct taxes we model include payments of
income tax due on labour income and benefit income, but not all of those that
are levied on other incomes (so we do not model the effect of the abolition of
dividend tax credits). We also model both employer and employee National
Insurance contributions. Employers’ National Insurance is sometimes thought
of as a tax on businesses, but, like income tax and employee contributions, it is
levied on wages and salaries, and so it is consistent with our methodology to
include it as a tax on labour income (see also Section 9.4). Reforms to the
benefit system are also analysed,5 and the expenditure taxes that we model
include VAT and most excise duties, but not stamp duty on house purchases.
Table 9.1 indicated that amongst the changes we analyse, reforms to income
tax, National Insurance and excise duties (notably including road fuel duty) all
had significant revenue effects. Many households also benefited from
increased transfers via benefits or credits and tax credits. The net modelled
effect of the changes considered here is to increase household incomes6 and
therefore loosen the government’s fiscal position next year by £9.8 billion,7
which is approximately £8 per household per week.8 Given that we calculated
the overall net effect of Labour’s reforms to be a fiscal loosening next year of
just £1.6 billion – nearer £1.50 per household per week – it is clear that by
focusing only on the subset of measures that directly affect households, we
show a much more generous average ‘giveaway’ than we would do if we
included all the tax and benefit changes. For example, we do not capture the
5
With the exception of some reforms to bereavement and incapacity benefits, which we
cannot model due to data limitations.
6
Net of direct taxes and benefits, credits and tax credits.
7
Cash changes are expressed in October 2002 prices.
8
The modelling that underlies these numbers and Figure 9.1 compares a world with the tax
system that will exist once all the reforms announced by January 2003 have been enacted, and
a hypothetical world in which the 1996–97 tax system had been retained and benefit rates and
expenditure taxes had been uprated in line with inflation whilst the parameters of the income
tax system were uprated in line with statutory rules. This means that we do include changes to
National Insurance and the new credits and tax credits that will come into force during 2003.
115
Green Budget, January 2003
relatively large tax revenues from dividend tax changes or increases in stamp
duties, although neither do we capture the more modest cuts in corporation
tax. Another reason for the discrepancy in the estimated budgetary effects is
that payments of (for example) expenditure taxes that are levied on institutions
other than households will not show up in the analysis presented here.9
Figure 9.1 shows how the gains from our relatively generous subset of tax and
benefit changes are spread across the income distribution. In the figure, the
population of households is ranked according to income and then split into 10
equally sized groups or ‘deciles’. For each decile, the figure plots the
estimated average percentage change in net incomes resulting from the
reforms modelled.
Figure 9.1. Impact of direct personal tax, benefit and expenditure tax
changes since 1997
Percentage change in income
20
15
10
5
0
-5
Poorest
2
3
4
5
6
7
8
9
Richest
Income deciles
Notes: Deciles are constructed by ranking households according to their income – measured
before housing costs and adjusted for family size – and then splitting the population into 10
equally sized groups. The bars show the modelled impact on real incomes (measured in
October 2002 prices) of the majority of changes to direct and indirect taxes and to benefits and
credits that have occurred since 1997.
Source: Authors’ calculations based on results from the IFS tax and benefit model, TAXBEN,
run using Family Resources Survey 2000–01 and Family Expenditure Survey 1999–2000.
We can see that the reforms are progressive. Whilst net incomes across the
population increase by around 2% on average (shown by the black line),
incomes in the poorest group increase by slightly more than 15%. Increases
get steadily smaller as income rises and are almost exactly zero in decile 8.
There are small losses in decile 9, and the richest group experiences a loss of
almost 3% of its net income. In cash, rather than proportional, terms, average
gains are found to be £20 or more per week in each of the bottom three
9
In part, the discrepancy may also be due to undercounting of revenues from some
expenditure taxes in our model because the data available to us tend to under-report some
types of household expenditures.
116
Distributional effects of fiscal reforms since 1997
deciles, whilst average losses are around £6 per week in decile 9 and around
£28 per week in the richest group.
As well as looking at how changes vary across the income distribution, we can
also look at which types of family have gained from the reforms that we
model. This is of interest because, as well as targeting benefit increases and
tax cuts towards those with low incomes, Labour has also given special
priority to reducing poverty among children and among pensioners.10 It is not
surprising to find that pensioners and low-income families with children were
the biggest gainers from the package of reforms that we model. No-earner
couples with children are found to gain around £40 a week, on average, and
the figure is around £28 for single parents. Pensioner households also gain
more than £20 per week, on average. For groups containing working-age
adults with no children, average changes were of a smaller magnitude than the
gains amongst the elderly and those with children.
9.3 Limitations of the distributional
analysis
The above analysis measures the distributional impact of changes to taxes and
benefits where the payment can relatively easily be allocated to households
according to the personal incomes or expenditures of household members. It
therefore encompasses the majority of those taxes that are levied on the
earnings or spending of private individuals, plus the majority of transfer
payments made through the benefit system.
The ‘tax payment’11 that we allocate to each household is the sum of the tax
levied on all spending by household members and that levied on the incomes
received for the labour supplied by household members, minus any benefit
payments that they receive. To make comparisons between the distributional
effects of the tax system as it existed at different points in time, we assume
that (gross) incomes and expenditures (inclusive of tax payments) are not
affected by the tax system. In other words, for each household in the data, we
compare the tax payment implied by each tax system, given their observed
expenditure and labour income.
Using this notion of ‘tax payment’ is not the same thing as considering who
makes the cash transfer to the state: much of income tax and VAT is
administered via cash transfers between companies and the relevant authority.
Nor does it capture the economic notion of effective incidence – in other
words, who is ultimately made financially worse off as a result of a particular
tax being imposed rather than collecting the same revenue (and financing the
10
The different ways that Labour has used social security and direct tax reforms to target
support towards particular groups are discussed in M. Brewer, T. Clark and M. Wakefield,
‘Social security in the UK under new Labour: what did the third way mean for welfare
reform?’, Fiscal Studies, 2002, vol. 23, pp. 505–37 (see especially section VI).
11
This terminology and many of the ideas here are due to A. Dilnot, J. Kay and M. Keen,
‘Allocating taxes to households: a methodology’, Oxford Economic Papers, 1990, vol. 42, pp.
210–30.
117
Green Budget, January 2003
same pattern of government expenditure) from some other source. The
effective incidence of a tax depends on how prices charged and quantities
traded are affected by this tax, and we do not model these. Since the welfare of
households will also ultimately depend on the amount that they consume and
the number of hours that they work, we also do not capture the welfare effects
of tax changes.
Figure 9.2. Impact of direct personal tax, benefit and expenditure tax
changes since 1997
Percentage change in income
20
Base case
15
Extra allocated as constant proportion
10
5
0
-5
-10
Poorest
2
3
4
5
6
7
8
9
Richest
Income decile
Notes: See notes to Figure 9.1.
Source: Authors’ calculations based on the numbers in Table 9.1 (see sources there) and
results from the IFS tax and benefit model, TAXBEN, run using Family Resources Survey
2000–01 and Family Expenditure Survey 1999–2000.
Nonetheless, our methodology does yield a first approximation to how the tax
system affects households. Accepting the usefulness of such an
approximation, the biggest weakness in the analysis is that, while we have
acknowledged that all taxes and benefits are ultimately paid to or by people,
we also omit a large set of reforms. We estimated that the reforms that we are
able to model would increase household incomes in 2003–04 by around
£8.2 billion more than will the full set of tax and benefit reforms since 1997. A
crude way to generate a comprehensive picture of the distributional
implications of all of these tax and benefit reforms would be to assume that
those changes that were omitted had an equal proportionate impact on all
households, reducing disposable incomes by approximately 1.7%.12 This is
equivalent to taking £2.25 per week, on average, from households in the
poorest decile and almost £17 per week, on average, from those in the richest
tenth. Allocating extra changes in this proportional way will not affect the
progressive shape of the distribution of proportional changes in incomes due
12
In allocating the full £8.2 billion to the UK household sector, we ignore the fact that some
taxes may be levied on the incomes or expenditures of individuals who are not UK residents.
118
Distributional effects of fiscal reforms since 1997
to the reforms modelled.13 This is shown in Figure 9.2, which compares the
results of Figure 9.1 with the results as they would look if the package
modelled did take an extra 1.7% from the incomes of all households.
The assumption of equal proportional changes is unlikely to reflect accurately
who would actually pay extra tax due to the measures omitted in Section 9.2.
An important set of taxes that were omitted are those levied on incomes that
derive from company profits or from the ownership of property. For example,
tax-raising changes to stamp duties and dividend tax credits, and cuts in
corporation tax, were not modelled. It is unlikely that these taxes have a
constant proportional impact on incomes across the income distribution. Their
omission was largely due to inadequate data rather than due to these taxes not
being amenable to analysis within the framework of ‘tax payments’. To help
clarify the conceptual framework, and to see how misleading the benchmark
case of a constant proportional impact is, in the next two sections we consider
‘taxes on business’ and stamp duty on residential property transactions.
9.4 Taxes on business
How much have taxes on business increased? The answer depends, to a large
extent, on what we think should be included under the heading of ‘taxes on
business’. At one extreme, we might start with the idea of all taxes paid to the
government by firms. This would include not only corporation tax and
business rates, but also most of VAT, all of income tax collected under the
PAYE system and all National Insurance contributions levied on the earnings
of employed workers, regardless of whether these are labelled as employer or
employee contributions. Most observers would agree that this definition would
be much too broad. At the other extreme, we might classify taxes according to
who feels their ‘effective incidence’, or who is ultimately made worse off by
their imposition. Unfortunately, this would lead us to the conclusion that there
are no taxes on business, as companies have a legal identity only and cannot
be made worse off in any meaningful sense.
A reasonable intermediate position might be to consider those taxes whose
effective incidence falls on the owners of companies. This does not produce a
neat, operational classification of taxes into those on business and those not on
business, and may suggest rather surprising results in some cases. For
example, in a small, perfectly competitive, open economy with a high degree
of international capital mobility but limited international mobility of labour, it
can reasonably be argued that much of the effective incidence of a sourcebased corporate income tax will ultimately fall on domestic workers rather
than on shareholders. With mobile capital, investors – regardless of where
they live – will only finance investment projects that generate a given post-tax
rate of return – regardless of where the activity is located. If a country imposes
a corporation tax that reduces the post-tax return on investment located in its
territory, the result will be that investors finance fewer investment projects in
13
Indeed, allocating changes in this fashion is equivalent to shifting the axis marking a zero
proportional change up by 1.7 percentage points on the bars showing the changes for each
decile.
119
Green Budget, January 2003
that country – only those projects with a pre-tax return high enough to pay the
required post-tax return after corporation tax will continue to be viable.
Workers in that country will then operate with less capital per worker, be less
productive and earn lower wages as a result. Assuming that migration is
insufficient to equalise wages across countries, it is workers rather than
shareholders in the country imposing the corporation tax who will ultimately
be worse off.14
Interesting as this may be, it does not accord well with popular perceptions of
what are considered as taxes on business. A more pragmatic and more
common approach focuses on rights to the sources of income on which
different taxes are levied. This is very similar to the way in which we allocated
tax payments in Section 9.2. There we allocated the payment of labour taxes to
the individual on whose labour income the tax is levied. A natural extension
would be to allocate taxes on incomes derived from company profits to the
owners of the company who have the right to the income stream on which the
tax is levied. Applying this idea suggests that taxes that are levied on company
profits, or on some component of company profits (e.g. dividends), can be
classed as a category of ‘taxes on (the owners of) business’. This may be
considered too narrow, but it is unclear what general principles would lead to a
satisfactory, broader definition.
Following this approach, corporation tax, which is levied on company profits
that are ultimately paid to the owners of companies, is classified as a business
tax. Income tax on wages and salaries is not classified as a tax on business,
since wages and salaries are paid to workers, not to shareholders. Income tax
on company dividends is however classed as a business tax, since dividends
are distributed profits, and are clearly paid to the owners of companies.
By the same principle, National Insurance contributions are classed as a tax on
individuals, regardless of whether these are nominally employer or employee
contributions. Like income tax, all National Insurance contributions are levied
on a base of wages and salaries (albeit with different rates structures), and it is
hard to think of any reasonable principle that would result in employer and
employee National Insurance contributions being treated differently. There
may be disagreement as to whether their effective incidence is mainly on
workers or mainly on the owners of firms, but a basic principle of public
economics is that the effective incidence of a tax should not depend, at least in
the long run, on whether it is levied on the buyer or the seller of a good or
service. Thus the effective incidence of employer National Insurance
contributions may be on employers, to the extent that it is not shifted onto
workers in the form of lower wages. But if this is so, the effective incidence of
employee National Insurance contributions would also fall on employers, as
higher wages would then need to be paid to attract the same workforce. In this
case, the effective incidence argument would quickly lead to personal income
14
Shareholders continue to earn the same post-tax rate of return on their investments, though
the companies they own may be locating less activity at home and more activity overseas.
Under these conditions, owners of companies can only be made worse off in the long run by
residence-based taxes levied on their worldwide investment income from all sources. Since
relatively little revenue is collected from such taxes, this approach would not classify most
taxes as ‘business taxes’.
120
Distributional effects of fiscal reforms since 1997
tax on wages and salaries also being classed as a business tax, which does not
seem a satisfactory result.
By how much have ‘business taxes’ changed since 1997?
Notwithstanding the limitations of this approach, we focus here on the extent
to which taxes levied on company profits have increased under Labour
governments since 1997. A broader approach might also include taxes levied
on some components of business expenditure – for example, business rates
levied on the occupation of non-domestic property and environmental taxes
such as the climate change levy and the aggregates tax. As it happens,
extending our analysis to include these taxes would have little impact on the
discussion in this section. Business rates have generally been increased in line
with inflation under Labour, as they were under the previous Conservative
administration. The introduction of the climate change levy and the aggregates
tax was intended to be revenue-neutral for the business sector overall, with a
corresponding reduction in employer National Insurance contributions.15
Even restricting our attention to taxes on company profits, there has been no
shortage of tax changes since 1997, with somewhat offsetting effects on
overall tax revenues. The principal changes are summarised in Table 9.2. The
biggest single increase was brought about by the abolition of repayable tax
credits on dividend income for pension funds and some other tax-exempt
shareholders, introduced in July 1997. This was, in effect, an increase in tax
levied on distributed profits for those shareholders. At the time, this was
estimated to raise upwards of £5 billion per annum for the exchequer by 1999–
2000, although, so far as we are aware, estimates for later tax years have not
been published.16
The second major increase was temporary, resulting from the introduction of
quarterly instalment payments of corporation tax for large companies. The
effect of the new system was to accelerate tax payments compared with the
old system of advance corporation tax (ACT). This brought in substantial
revenues over the four financial years up to 2002–03: it was estimated to
increase corporation tax receipts by, on average, around £2 billion in each of
those years. It should be noted that this reform has a major effect on tax
revenues if figures are presented on a cumulative basis for the period from
1997 onwards, but has little impact if figures are presented for individual years
from 2003–04 onwards.17 Immediately prior to this temporary rise in
15
Note that the offsetting reduction in employer NICs would not be classed as a reduction in
business taxes, following the general approach of this section. However, the extent to which
the effective incidence of these environmental taxes falls on the owners of companies, or is
shifted onto consumers in the form of higher product prices, is also unclear. Without further
analysis of such issues, it would seem unduly harsh to treat this revenue-neutral policy as
imposing an increase in taxes on business.
16
Our estimates for later years simply uprate the initial £5.4 billion in line with nominal GDP
growth.
17
Actually, there is a small negative effect of this reform from 2003–04 onwards. This comes
mainly from the loss of revenue from ‘surplus ACT’. ACT only affected the timing of tax
payments for most firms, but had the effect of increasing total tax payments for firms in a
121
Green Budget, January 2003
corporation tax payments, there was a temporary contribution to government
revenues of a similar amount in 1997–98 and 1998–99 from the windfall levy
imposed on privatised utilities. More recently, there is a small increase in tax
revenue resulting from changes to the taxation of North Sea oil and gas
production.
Table 9.2. Revenue effects of major changes in ‘business taxes’ (£ billion)
1997–98
1998–99 1999–2000 2000–01
2001–02
2002–03
2003–04
Corporation tax rate cuts:
July 1997
–1.6
–2.2
–2.3
–2.4
–2.5
–2.6
March 1998
–0.8
–1.1
–1.2
–1.2
March 1999
–0.1
–0.1
–0.1
April 2002
–0.0
–0.3
Total
–1.6
–2.2
–3.1
–3.6
–3.8
–4.3a
R&D tax credits
–0.1
–0.5
–0.6
ACT replaced by
0.1
1.6
2.0
3.1
2.2
–0.5
quarterly instalments
Dividend tax credits
2.3
4.0
5.4
5.7
5.9
6.2
6.5
North Sea taxation:
Introduction of 10%
0.1
0.5
supplementary charge
and 100% capital
allowances
Abolition of licence
royalties
–0.2
Windfall levy
2.6
2.6
Total revenue effect
4.9
5.1
4.8
4.6
5.3
4.1
1.4a
a
Differences between these figures and those in Table 9.1 arise from the fact that here we only
consider reforms that were announced by Labour after the party was elected in 1997, whereas
Table 9.1 includes costings for all measures enacted since 1997. Table 9.1 is also more
comprehensive and includes the estimated effects of a number of technical changes and antiavoidance measures that are excluded from this table of major reforms.
Notes: Tax rate cuts include the revenue effects of cuts in the small companies’ rate and the
introduction and subsequent reduction in the starting rate. Initial costings come from Budget
and Pre-Budget Reports, and where necessary these have been uprated to later years using the
index of nominal GDP growth published by the Office for National Statistics and (for
projections) HM Treasury.
Sources: HM Treasury, Financial Statement and Budget Report, various years; HM Treasury,
Pre-Budget Report, 2002; Inland Revenue Press Release 2, 2 July 1997; Inland Revenue Press
Release 8, 17 March 1998; Inland Revenue Press Release 9, 17 March 1998.
These increases in taxes on company profits have been offset by cuts in the
main rate of corporation tax, which was reduced from 33% to 31% in the July
1997 Budget at a cost of around £2 billion per annum, and to 30% in the
March 1998 Budget at a cost of around £1 billion per annum. Corresponding
reductions in the small companies’ corporation tax rate are costing the
exchequer around £0.5 billion per annum. Further tax cuts include the
introduction of a 10% starting rate of corporation tax in the March 1999
Budget, which was reduced to 0% in the April 2002 Budget, and the
introduction of new tax credits for research and development.
‘surplus’ ACT position, which could not fully offset their ACT payments against mainstream
corporation tax liabilities.
122
Distributional effects of fiscal reforms since 1997
An accurate estimate of the overall effect of these changes is difficult to
produce, as official costings of different reforms have covered different time
periods and some are now rather outdated. A ballpark figure for 2002–03
would suggest an increase in the region of £4 billion, reflecting around
£8 billion extra revenue from dividend taxation and quarterly instalments,
offset by around £4 billion from rate cuts and other changes. For 2003–04, we
estimate a smaller increase, in the region of £1–£2 billion, as the effect of the
transition to quarterly payments is no longer present. Much then depends on
the particular year for which any estimates are reported, and, as noted earlier,
the effect of the switch from ACT to quarterly instalments appears much more
significant if measured cumulatively over this period rather than for the final
year that we consider. Our discussion here also excludes the effect of a host of
technical changes and anti-avoidance measures,18 the costs of which are
inherently difficult to estimate accurately, and a number of measures limited to
smaller firms.19
We conclude by reiterating that there is no compelling way of classifying
particular taxes as ‘taxes on business’. In our distributional analysis, we have
followed an approach that allocates ‘tax payments’ to the individual who has
the right to the income stream on which a tax is levied, or whose resources are
used to fund an expenditure on which a tax is levied. Largely due to
limitations of data, taxes levied on incomes derived from company profits
were not allocated to households.
These taxes on company profits have been higher over the period 1997–98 to
2002–03 as a result of changes introduced by Gordon Brown. The
government’s emphasis on changes to corporation tax rates does not tell the
full story here. On the other hand, estimates of the scale of the increase in
‘business taxes’ can easily be exaggerated. Moreover, some of these effects
are explicitly temporary, and it is expected that revenues from these taxes in
2003–04 will be only around £1–£2 billion greater than would have been the
case had Labour not introduced any of these reforms. This is small compared
with the net giveaway of £9.8 billion modelled in Section 9.2, and so
allocating the payment of ‘business taxes’ to individuals according to their
ownership shares in companies would not significantly affect the distributional
results found there. Further, since share ownership tends to be concentrated in
the upper reaches of the income distribution,20 allocating these taxes might add
to the progressive overall effect of reforms since 1997.
18
Examples include stricter limits on the carry-back of losses, changes to the taxation of
insurance companies and changes to the rules affecting controlled foreign companies.
19
Examples include higher capital allowances for investment in plant and machinery, and tax
relief for venture capital trusts.
20
The pattern of share ownership in the UK, and how little data we have on the amount of
wealth individuals have in pension funds, are discussed in J. Banks and M. Wakefield,
‘Stockholding in the United Kingdom’, chapter 8 of L. Guiso, M. Haliassos and T. Jappelli
(eds), Stockholding in Europe, Palgrave Macmillan, Hampshire, 2003.
123
Green Budget, January 2003
9.5 Stamp duty on residential properties
Although stamp duty is paid on transactions of both residential and nonresidential properties, here we only consider sales of residential properties. We
restrict our focus in this way because it is easier to allocate payments within
the income distribution when the buyer is an individual or a family, rather than
a company.
In the analysis of Section 9.2, revenues are deemed to have been raised from
expenditure taxes if there have been increases in the rate of the tax or
expansion of the tax base (i.e. the total value of the transactions on which the
tax is levied). Considered in this way, stamp duty on residential properties
would have raised money since 1997 for two reasons: first, the threshold
house price at which tax payments begin has been fixed in nominal terms,
which means that some revenues have been due to the fact that house price
inflation has moved more property transactions into the stamp duty tax base;
secondly, a graduated structure has been added to what was previously a tax
with only one positive tax rate. In this section, we discuss only the revenues
raised from the introduction of the graduated structure.21
When Labour came to power in 1997, stamp duty was paid at a rate of 1% on
the value of property transactions that exceeded £60,000. Transactions of
£60,000 or less were not liable for the tax. Since 1997, a graduated structure
has been introduced via a series of reforms, as shown in Table 9.3.
Table 9.3. Rate of stamp duty on property, 1997 to present day
Transaction
value (£000)
0–60
60–250
250–500
500+
Before
8 July 1997
8 July 1997
– 23 March
1998
0%
1%
1%
1%
0%
1%
1.5%
2%
24 March
1998
– 15 March
1999
0%
1%
2%
3%
16 March
1999
– 27 March
2000
0%
1%
2.5%
3.5%
28 March
2000 to date
0%
1%
3%
4%
To summarise, the rate of stamp duty on any residential property that sells for
more than £250,000 has increased since 1997. It has increased from 1% to 3%
for properties that sell for between £250,001 and £500,000, and from 1% to
4% for properties that sell for more than £500,000.
Labour has not changed the fact that stamp duty operates as an average rate
tax. This means that if a property transaction falls into the top stamp-duty
band, then the tax is paid at 4% on the full value of the sale, not just the value
exceeding £500,000. Similarly, if a transaction falls into one of the lower
bands, then 1% or 3% tax will be levied on the entire transaction value. To
take an example, a £300,000 property transaction will be liable for £9,000 of
stamp duty, which is 3% of £300,000.
21
Although some of these revenues are themselves due to the fact that since the graduated
structure was first created in 1997, the thresholds for higher-rate bands have not been
increased to allow for inflation.
124
Distributional effects of fiscal reforms since 1997
The data in Table 9.1 indicate that £2.1 billion extra revenue has been raised
from stamp duties since 1997, but they do not isolate how much of this has
come from the introduction of a graduated system for residential properties.
We approximate for this figure by using data published by the Inland Revenue
that give the yield of stamp duty on residential properties by the price band of
property transactions. For example, in 2001–02, when the increases to 3% and
4% had been fully implemented, we know that yields were as given in the first
column of Table 9.4.
Table 9.4. Yield of stamp duty on residential property, 2001–02
Stamp duty band
Yield
£ million
Proportion due
Estimated £millions
to change in
due to change in rates
rates
965
⅔
643
£250,001–£500,000
715
¾
536
£500,001+
1,680
n/a
1,180
Total
Sources: Inland Revenue Statistics, www.inlandrevenue.gov.uk/stats; authors’ calculations.
If the structure of stamp duty had not been changed after 1996–97, then for
property transactions exceeding £60,000 in value, the tax yield would have
been 1% of transaction values. Due to changes in the tax structure, the tax
yield for transactions in the £250,001–£500,000 price range in 2001–02 was
actually 3% of transaction values. If we assume that property prices in 2001–
02 were not affected by the changes in the structure of stamp duty after
1997,22 then we can state that the extra stamp duty paid on property
transactions in this price range due to increases in the tax rate was 2% of
transaction values (the excess of 3% over 1%). In other words, two-thirds of
the tax yield for these transactions was due to the reforms. By similar
reasoning, we can argue that three-quarters of the yield on transactions of
£500,001 or more was due to increases in the tax rate. Applying this
reasoning, Table 9.4 shows that we estimate the amount of extra tax paid to be
around £1.18 billion. In other words, if we had included this change in stamp
duty in our distributional analysis in Section 9.2, then the total cost to the
exchequer of the reforms considered would have been around £8.6 billion (or
£7 per family per week, on average) rather than £9.8 billion (or around £8 per
family per week).
We now consider how payments of this stamp duty might be spread across the
income distribution. Our analysis of this issue will be conducted in a slightly
different way from that in which we allocated payments of expenditure taxes
in Section 9.2, but it might give a reasonable approximation for how we would
have allocated these duties. As we now explain, the results might also be of
interest if the financial cost of stamp duty actually falls on house-sellers rather
than house-buyers.
Throughout this chapter, in allocating tax payments we have assumed that
total expenditures remain unaltered after a reform and that the extra tax is paid
out of these expenditures. We then allocate the tax payment to the household
22
This is unlikely to be the case, as we would expect to observe spikes in the distribution of
sale values just below the thresholds.
125
Green Budget, January 2003
containing the person who made the particular expenditure (that is, to the
buyers of goods and services). In order to allocate payments of stamp duty on
properties in this way, we would need to know the incomes of house-buyers
and the amount that they paid for their new homes. But data of this kind are
not easily available. We can, though, observe the incomes and estimated
property values for homeowners in the British Household Panel Survey
(BHPS) for 2000. We will allocate the cost of stamp duty across the income
distribution according to these data.
To the extent that house-buyers have different characteristics from
homeowners, allocating payments of stamp duty according to the values of the
stock of properties owned and the characteristics of owners will give us
different results from allocating according to the values of properties that are
traded and the characteristics of buyers. First-time buyers are perhaps
especially likely to fall into different age and income groups from typical
homeowners. Also, the approximation will only be accurate to the extent to
which homeowners tend to stay within the same stamp-duty band when they
move home. The results of our method are also of interest in themselves
because they tell us who loses out from an increase in stamp duty if the tax
change results in a fall in the price that owners can expect to receive for their
properties.
Table 9.5. Distribution of the value of homes worth more than £250,000,
across the income distribution
Decile of total population income
Percentage of value of total stock of
distribution
homes worth £250,000+
Poorest
6.4
2
1.3
3
3.1
4
4.8
5
6.3
6
7.7
7
5.0
8
14.6
9
14.8
Richest
36.0
Total
100.0
Notes: Deciles are constructed by ranking households according to their income adjusted for
family size and then splitting the population into 10 equally sized groups. Also, see footnote
23.
Sources: British Household Panel Survey 2000; authors’ calculations.
For each income decile, we will allocate a proportion of the cost of the extra
stamp duty paid that is equal to the proportion of the total value of the stock of
all properties in each stamp-duty band that is owned by people in that income
group. For a sample of 4,824 households from the 2000 BHPS, Table 9.5
reports percentages of the value of the stock of all properties worth more than
£250,000 that were owned by households in each income decile. In the table,
we do not split properties according to whether or not they are worth more or
126
Distributional effects of fiscal reforms since 1997
less than the upper £500,000 stamp-duty threshold because there were too few
very valuable homes in the data to make such a split interesting.23
We have estimated that, in total, the changes to stamp duty that we are
considering raised around £1.18 billion for the exchequer, or slightly less than
£1 per household per week on average. Allocating the stamp duty payment in
line with the values of homes owned at different points in the income
distribution in our sample, we find that approximately £825 million of the
increase would be paid by the richest 30% of the population. Of this, almost
£450 million would be allocated to the richest tenth. This implies households
in the richest tenth paying extra tax worth approximately £3.50 per week (or
0.4% of their income), on average, due to these stamp duty reforms. This is
small compared with net losses of around £28 per week (or 3% of income) that
we estimated to be taken, on average, from those in the richest income group
by the package of reforms considered in Section 9.2.
Figure 9.3. Impact of direct personal tax, benefit and expenditure tax
changes since 1997, with and without effect of stamp duty
Percentage change in income
20
No stamp duty
15
With stamp duty
10
5
0
-5
Poorest
2
3
4
5
6
7
8
9
Richest
Decile
Notes: See Figure 9.1.
Source: Authors’ calculations using the British Household Panel Survey for 2000 and results
from the IFS tax and benefit model, TAXBEN, run using Family Resources Survey 2000–01
and Family Expenditure Survey 1999–2000.
The effect of stamp duty is small relative to the package considered in Section
9.2. Figure 9.3 shows that it also does not affect the progressive shape of the
results we found there. The figure reproduces the results of Figure 9.1 and
compares them with a set of results that include the distributional effects of
stamp duty as modelled here. If anything, the stamp duty reform tends to add
23
There were actually 187 households with homes worth more than £250,000, but only 15 of
these homes were valued at more than £500,000. We do use information on whether or not a
home falls into the top stamp-duty band when calculating how much extra tax households in
each decile would pay.
127
Green Budget, January 2003
to the progressive shape of the package since, as modelled (and with the
exception of a relatively large proportionate income loss in the poorest tenth),
it tends to take proportionally slightly more from those in higher income
deciles. Raising money from stamp duties is certainly progressive in the sense
that it takes the largest cash amounts from people towards the top of the
income distribution.
9.6 Conclusion
In this chapter, we have looked at the distributional impact of reforms to taxes
and benefits that have been implemented since 1997. Amongst the reforms we
model in Section 9.2, the pattern has generally been progressive. Many of the
reforms that we do not model are tax increases. This means that if we could
include the distributional impact of all these reforms, we would be assessing a
package that is less generous to households than the package considered in
Section 9.2. We have seen that attempting to model the reforms that were
omitted raises a series of difficult conceptual questions and problems of data
availability. Some of the taxes that are omitted are levied on incomes derived
from company profits or from owning property. It is likely that the payment of
these taxes falls relatively heavily on those high up the income distribution,
and so it seems safe to assume that, in a more comprehensive distributional
analysis, the progressive nature of the package of reforms made since 1997
would not be compromised.
Stephen Bond and Matthew Wakefield
128
Appendix A: Forecasting public
finances
This appendix describes the techniques used for our public finance forecasts. It
starts by comparing the forecasts made for borrowing in 2001–02 in last year’s
Green Budget and the November 2001 Pre-Budget Report with the eventual
out-turn. It then goes on to provide more background information to the shortterm and medium-term public finance forecasts that are set out in Chapter 3.
A.1 The accuracy of our previous forecasts
The November 2002 Pre-Budget Report1 gave an out-turn figure for a balance
of £0.0 billion on public sector net borrowing in 2001–02 (excluding spending
associated with the windfall tax). This was almost exactly halfway between
the £1.4 billion deficit expected by the Treasury in the November 2001 PreBudget Report2 and the £1.6 billion surplus forecast in the January 2002 IFS
Green Budget.3 Table A.1 shows both forecasts alongside the estimated outturn for 2001–02 from the November 2002 Pre-Budget Report.4
Both the Treasury and IFS forecast higher current receipts than were realised.
Both of these forecasts for receipts were reasonably accurate, being within 1%
of the final out-turn.
Table A.2 shows the breakdown of both the Treasury’s and IFS’s main errors
in forecasting tax receipts for 2001–02. Although the overall effect was an
overestimate of receipts, both sets of predictions underestimated both income
tax and council tax, with income tax presenting the greatest deviation in any
direction. The November 2001 Pre-Budget Report underestimated income tax
receipts by £6.0 billion, while the IFS forecast was £4.6 billion too low. The
most significant overestimates were for corporation tax and social security
1
HM Treasury, Pre-Budget Report 2002, Cm. 5664, London, 2002 (www.hmtreasury.gov.uk/pre_budget_report/prebud_pbr02/prebud_pbr02_index.cfm).
2
HM Treasury, Pre-Budget Report 2001, London, 2001 (www.hmtreasury.gov.uk/pre_budget_report/prebud_pbr01/prebud_pbr01_index.cfm).
3
A. Dilnot, C. Emmerson and H. Simpson (eds), The IFS Green Budget: January 2002,
Commentary no. 87, Institute for Fiscal Studies, London, 2002
(www.ifs.org.uk/gbfiles/gb2002.shtml).
4
More recent information on the out-turn for 2001–02 suggests that current receipts were
£390.4 billion, that total managed expenditure was £389.7 billion and that therefore there was
a surplus on public sector net borrowing (including spending associated with the windfall tax)
of £0.7 billion. This suggests that receipts were closer to the Treasury forecast, and that
spending and public sector net borrowing were closer to the January 2002 IFS Green Budget
forecast, than is suggested by the PBR out-turn figures discussed here. Source: Office for
National Statistics Press Release, ‘Public Sector Accounts: 3rd Quarter 2002’, 23 December
2002 (www.statistics.gov.uk/pdfdir/psa1202.pdf). In this appendix, we examine the PBR outturn rather than the latest available figures since no breakdown by type of expenditure or tax is
currently available for the latest figures.
129
Green Budget, January 2003
contributions. The November 2001 Pre-Budget Report overestimated their
yields by £0.9 billion and £1.1 billion respectively, while the IFS forecasts
were £0.9 billion and £0.3 billion too high respectively.
Table A.1. A comparison of last year’s IFS Green Budget forecast and the
Treasury November 2001 Pre-Budget Report forecast with the estimated
out-turn for 2001–02 from the November 2002 Pre-Budget Report (£
billion)
HM Treasury
Pre-Budget
Report forecast,
November 2001
391.2
393.7
IFS Green
Budget forecast,
January 2002
Current receipts
Total managed expenditure
Of which:
Departmental expenditure limits
212.5
Annually managed expenditure
181.1
PSNBa
1.4
a
PSNB excludes spending financed by the windfall tax.
391.4
391.1
Estimate,
Pre-Budget
Report,
November 2002
390.7
391.8
210.0
181.1
–1.6
212.1
179.7
0.0
Table A.2. IFS Green Budget and Treasury main errors in forecasting tax
receipts, 2001–02 (£ billion)
Pre-Budget Report
forecast,
November 2001
–6.0
0.9
0.3
0.3
1.1
–0.5
4.4
0.5
IFS Green Budget
forecast,
January 2002
–4.6
0.9
0.3
0.2
0.3
–0.5
4.1
0.7
Income taxa
Corporation tax
Value added tax
Fuel duties
Social security contributions
Council tax
Other taxes and receipts
Total
a
Net of tax credits.
Source: Out-turn figures for 2001–02 from HM Treasury, Pre-Budget Report 2002, Cm. 5664,
London, 2002 (www.hmtreasury.gov.uk/pre_budget_report/prebud_pbr02/prebud_pbr02_index.cfm).
A.2 Techniques used in our forecasts
For the current financial year, three different sources of information are
examined before coming to a judgement for each element of government
revenue. In addition to the latest Treasury forecast from the November 2002
Pre-Budget Report, we use information from the revenues implied by a current
receipts method, and the IFS modelled approach.5
5
For a more detailed explanation of both these techniques, see C. Giles and J. Hall,
‘Forecasting the PSBR outside government: the IFS perspective’, Fiscal Studies, 1998, vol.
19, pp. 83–100.
130
Appendix A
1. Information from current receipts. The current receipts method uses the
information on receipts received in the current financial year compared
with those received up to the same point in the previous financial year. An
estimate for the current year’s receipts is then calculated using the
following formula:
2002–03 forecast =
Receipts received so far this year
× 2001–02 receipts.
Receipts received to the same point last year
While this is useful when forecasting revenues in the current financial
year, it cannot provide projections for borrowing in future years. Also,
caution should be used when revenues are cyclical or changes have been
made that may affect the timing of payments – for example, the effect of
moving to a quarterly system of corporation tax payments.
2. The IFS modelled receipts approach. This estimates growth in each of
the taxes using forecasts for the growth in the tax base relevant to each tax,
combined with an estimate of the elasticity of revenue with respect to the
growth in the tax base. Information on the revenue effects of preannounced tax changes from previous Budgets is then added in order to
reach a forecast. Hence, modelled receipts can be summarised by the
following formula:
2002–03 forecast = (2001–02 receipts × Tax-base change × Elasticity) + Tax changes.
This technique enables forecasts to be made for future years, given the
expected structure of the tax system. It should be noted that these forecasts
become considerably less accurate for later years, since forecasts for
changes in tax bases, estimates of elasticities and the impact of tax changes
all become less accurate.
The elasticities are largely estimated from TAXBEN, the IFS tax and
benefit model. The estimates for income tax elasticities are supplemented
by a model of the responsiveness of income tax revenues to changes in
employment and wages. For fuel, an elasticity calculated from previous
IFS research is used.6 Elasticities for beer, spirit, wine and tobacco duties
are taken from the median elasticity found in a range of UK studies.7
A.3 Forecasts for 2002–03
The Green Budget forecast is a judgement based on the Treasury’s latest
forecast contained in the November 2002 Pre-Budget Report, the current
receipts method and the IFS modelled approach. Each of these is presented in
Table A.3. Overall, we expect lower receipts than the Treasury, leading to
higher borrowing and a larger deficit on current budget.
6
L. Blow and I. Crawford, The Distributional Effects of Taxes on Private Motoring,
Commentary no. 65, IFS, London, 1997.
7
M. Chambers, ‘Consumers’ demand and excise duty receipts equations for alcohol, tobacco,
petrol and derv’, Government Economic Service, Working Paper no. 138, August 1999.
131
Green Budget, January 2003
Table A.3. Forecasts for government borrowing in 2002–03 (£ billion)
Pre-Budget
Report
Nov. 2002
Current
receipts
IFS
forecasting
model
IFS
forecast
judgement
Inland Revenue
Income tax (gross of tax credits)
114.1
111.9h,i
111.9h
110.0h
a
h
h
Corporation tax (CT)
29.3
28.3
33.4
28.5h
b
Tax credits
–3.5
n/a
n/a
n/a
Petroleum revenue tax
1.1
0.8
1.3
1.1
Capital gains tax
2.0
n/a
3.1
2.0
Inheritance tax
2.4
2.4
2.4
2.4
Stamp duties
8.2
7.6
7.6
7.8
Social security contributions
65.5
63.9
64.9
65.0
Total Inland Revenue (net of tax credits)
219.1
214.9
224.5
216.8
Customs and Excise
Value added tax (VAT)
64.5
63.7
64.5
63.7
Fuel duties
22.4
22.2
22.5
22.4
Tobacco duties
8.2
8.4
8.1
8.2
Spirit duties
2.2
2.3
2.2
2.2
Wine duties
1.9
2.0
2.1
1.9
Beer and cider duties
3.1
3.2
3.2
3.1
Betting and gaming duties
1.3
1.2
1.4
1.3
Air passenger duty
0.8
0.8
0.8
0.8
Insurance premium tax
2.1
2.2
2.0
2.1
Landfill tax
0.5
0.5
0.6
0.5
Climate change levy
0.9
0.9
0.9
0.9
Aggregates levy
0.2
0.2
0.2
0.2
Customs duties and levies
2.0
1.9
2.0
2.0
Total Customs and Excise
110.1
109.4
110.5
109.3
Vehicle excise duties
4.4
4.4
4.3
4.4
Oil royalties
0.5
0.5
0.5
0.5
Business ratesc
18.0
18.0
18.3
18.0
Council tax
16.6
16.6
16.0
16.6
Other taxes and royaltiesd
10.9
10.9
10.4
10.9
Total taxes and social security contribnse
379.6
374.8
384.5
376.5
Accruals adjustments on taxes
–0.6
–0.6
–0.6
–0.6
Less Own resources contribution to EU
–3.0
–3.0
–3.0
–3.0
Less Public corporations’ CT payments
–0.2
–0.2
–0.2
–0.2
Tax creditsf
1.2
1.2
1.2
1.2
Interest and dividends
4.1
4.1
4.1
4.1
Other receipts
18.6
18.6
18.6
18.6
Current receipts
399.7
394.9
404.6
396.6
Current spending
405.5
405.4
405.4
405.4
Current balanceg
–5.7
–10.6
–0.8
–8.8
Net investment
14.3
13.3
13.3
13.3
Public sector net borrowingg
20.1
23.9
14.1
22.1
a
National accounts measure: gross of enhanced and payable tax credits. b Includes enhanced
and payable company tax credits. c Includes district council rates in Northern Ireland.
d
Includes money paid into the National Lottery Distribution Fund. e Includes VAT and
‘traditional own resources’ contributions to EC budget. Cash basis. f Excludes children’s tax
credit and other tax credits that score as a tax repayment in the National Accounts. g Includes
expenditure associated with the windfall tax. h Net of tax credits. i Includes capital gains tax.
Sources: Treasury forecasts from HM Treasury, Pre-Budget Report 2002, Cm. 5664, London,
2002 (www.hm-treasury.gov.uk/pre_budget_report/prebud_pbr02/prebud_pbr02_index.cfm) –
this table is similar to table B12 on page 197); information on current receipts from ONS /
HM Treasury, ‘Public Sector Finances: December 2002’, Press Release, 21 January 2003
(www.statistics.gov.uk/pdfdir/psf0103.pdf); IFS calculations.
132
Appendix A
Inland Revenue receipts
For income tax (net of tax credits), we forecast £110.0 billion, which is lower
than the PBR forecast, since the Treasury forecasts gross receipts of
£114.1 billion with tax credits accounting for around £3 billion. Our estimate
takes into account the fact that the current receipts estimate is for
£109.9 billion (assuming capital gains tax of £2.0 billion), but allows for a
slight improvement in the strength of income tax receipts towards the end of
the financial year.
Our forecast for corporation tax is £28.5 billion, which is £0.8 billion below
the Treasury forecast of £29.3 billion. This is due to the current receipts
forecast suggesting that receipts will be just £28.3 billion. No weight is given
to the IFS modelled forecast since this does not take into account any decline
in company profits associated with the fall in the stock market. While the
current receipts forecast for corporate tax receipts has not been useful in the
past due to the change in timing of corporation tax receipts, our calculations
suggest that this should no longer be a problem, as there is no longer a timingof-receipts issue arising from the transition from the old corporation tax
system to the new one.
Our forecast for stamp duties is receipts of £7.8 billion this year. This is
slightly higher than the current receipts forecast and the IFS modelled forecast,
both of which are for £7.6 billion of receipts, but still £0.4 billion below the
PBR forecast of £8.2 billion.
With social security contributions, we forecast £65.0 billion, which is
£0.5 billion below the Treasury’s forecast of £65.5 billion. This is due to
evidence from both the IFS forecasting model (£64.9 billion) and the current
receipts method (£63.9 billion), which suggests that the Treasury’s forecast
might not be met.
Customs and Excise taxes
We forecast VAT receipts of £63.7 billion, lower than the Treasury’s forecast
of £64.5 billion. This is due to the information from current receipts, which
suggests revenues will be lower than the Treasury’s forecast.
We forecast that fuel duties will yield £22.4 billion, which is the same as the
Treasury’s forecast. This is because both current receipts and the IFS
forecasting model give very similar forecasts.
Other government receipts
For all other receipts, we take the Treasury’s forecast.
Government expenditure
We forecast current spending to equal £405.4 billion, which is just
£0.1 billion lower than the Treasury’s forecast. This is because we assume that
the £0.1 billion left in the AME (annually managed expenditure) margin from
the Pre-Budget Report is not used. We also assume that the special reserve
addition of £1 billion, which has been allocated to finance the potential
conflict in Iraq, is spent in 2002–03. While some or all of this might be carried
forward into 2003–04, the timing of the expenditure will not have any direct
133
Green Budget, January 2003
impact on the public finances or the interpretation of whether the fiscal rules
have been breached.
Our forecasts allow for a £1 billion underspend on public sector net
investment in 2002–03. This is due to the growth in public sector net
investment seen so far this year being lower than that forecast by the Treasury
for the year as a whole.
Government borrowing
As a result of lower government revenues (offset very slightly by lower
current expenditure), we forecast a deficit on current budget of £8.8 billion
for 2002–03. This is some £3.1 billion lower than the £5.7 billion forecast by
the Treasury.
We forecast that public sector net borrowing will be £22.1 billion, which is
£2.0 billion more than the £20.1 billion forecast by the Treasury.
A.4 Medium-term forecasts
Any assessment of the fiscal stance, and whether the Chancellor is going to be
successful in meeting his two fiscal ‘rules’, should be judged over the
economic cycle. Table A.4 presents the macroeconomic forecasts underlying
the baseline IFS forecast for government borrowing. For our central forecast,
we use the Treasury’s ‘cautious’ forecast for GDP growth of 2% in 2002–03,
2¾% in 2003–04, 3% in 2004–05 and 2¾% in 2005–06. In 2006–07 and
2007–08, we expect growth to be in line with the Treasury’s lower-bound
forecast for trend growth in those years of 2½% and 2¼% respectively.
Table A.4. Main macroeconomic assumptions used in the baseline
forecast (% growth in variable)
2002–03 2003–04 2004–05 2005–06 2006–07 2007–08
Gross domestic product (GDP)
2
2¾
3
2¾
2½
2¼
Real consumers’ expenditure
3¼
2½
2¾
2¼
2¼
2
Employment
0
¼
¼
¼
¼
¼
Real wages
2
2½
2¾
2½
2¼
2
GDP deflator
2½
2¼
2½
2½
2½
2½
Sources: GDP forecasts and GDP deflator from table B3 of HM Treasury, Pre-Budget Report
2002, Cm. 5664, London, 2002 (www.hmtreasury.gov.uk/pre_budget_report/prebud_pbr02/prebud_pbr02_index.cfm). Consumers’
expenditure based on table A8 of HM Treasury, Pre-Budget Report 2002, until 2004–05;
following GDP growth thereafter. Employment growth: IFS estimates consistent with table
A3 of HM Treasury, Pre-Budget Report 2002. Real wages growth: IFS estimates based on
GDP growth.
The IFS model also requires forecasts for growth in consumers’ expenditure,
employment and wages. These are set out in Table A.4 too. Where possible,
published Treasury forecasts are used. The model also requires a forecast of
growth in corporate profits. Due to the difficulties in forecasting this,
particularly in the present climate, we instead assume that, in the medium
term, corporation tax receipts rise back to the average level seen in recent
years.
134
Appendix B: Distributional effects of
pre-announced direct tax and
benefit reforms due in 2003–04
This appendix considers the distributional impact of reforms to direct taxes
and benefits that were announced in Pre-Budget Report 2001 and Budget 2002
but that do not take effect until 2003–04. The fact that these reforms will have
a significant effect on the incomes of many families at the beginning of the
new financial year in April (with further changes affecting those aged 65 or
over in October) will doubtless be in the Chancellor’s mind as he decides what
measures to announce in Budget 2003 and when they should take effect.
Our analysis shows that just over 60% of the money raised by the tax rises due
in the coming financial year will be offset by extra spending on benefits and
tax credits, leaving a net increase in personal taxes less benefits of around
£3 billion a year.1 On average, those aged 60 or over gain from the increases
in the minimum income guarantee (MIG), the basic state pension and the
pension credit and they pay little extra tax. Families without children pay more
tax with few compensating gains, while families with children lie somewhere
in between. The overall impact will be progressive, with families gaining, on
average, at the bottom of the income distribution and losing, on average, at the
top. This is in line with the impact of Labour’s other changes to direct taxes
and benefits since 1997.2
B.1 What reforms do we consider?
The complete list of tax and benefit reforms taking effect in 2003–04 can be
derived from the 2002 Budget.3 Those considered in this analysis are listed
below. All are effective from April 2003, except the pension credit.4
1
All figures in this appendix are in 2002 prices. Our estimates are that the tax rises will raise
£7.9 billion, that the benefit and credit changes in April 2003 will cost £3.8 billion and that the
pension credit will have a full-year cost of £2.1 billion. However, the pension credit starts in
October 2003, halving its estimated cost in 2003–04; if we counted the full-year cost of the
pension credit, then the changes in 2003–04 would only raise £2 billion.
2
See Chapter 9.
3
Chapter A of HM Treasury, Financial Statement and Budget Report, HC592, Stationery
Office, London, 2002 (www.hm-treasury.gov.uk/budget/bud_index.cfm). Benefit rates taking
effect from April 2003 were confirmed in Department for Work and Pensions, ‘Pensioners
and New Mothers Set to Gain from Cash Gains’, Press Release, 18 November 2002,
www.dwp.gov.uk/mediacentre/pressreleases/2002/nov/cfd1811-uprt.htm.
4
There are other, more minor, changes to benefits and personal taxes due which are not
considered in our model, such as increases in statutory maternity pay and the maternity
allowance. Taxes and benefits that have been increased in line with inflation do not count as
changes. Chapter 5 analyses the changes to National Insurance; HM Treasury, The Child and
135
Green Budget, January 2003
•
A freeze in the personal allowance for income tax for those aged under
65, equivalent to a real reduction of 1.7%.5
•
A freeze in the thresholds beyond which employers, employees and the
self-employed pay National Insurance (NI) contributions, again
equivalent to a real reduction of 1.7%.
•
A 1 percentage point increase in employee NI contributions, paid on all
earnings above the primary threshold.
•
A 1 percentage point increase in employer NI contributions, paid on all
earnings above the secondary threshold.
•
A 1 percentage point increase in self-employed NI contributions, paid
on all profits above the lower profits limit.
•
Overindexation of the personal allowance for those aged 65 or over.
•
Introduction of the child tax credit and the working tax credit, and
abolition of the children’s tax credit and the working families’ tax
credit.6
•
Abolition of child dependency increases in non-means-tested benefits.7
•
An above-inflation increase in the basic state pension for pensioners
(and linked benefits), and an increase in the MIG for those aged 60 or
over.
•
Introduction of the pension credit for those aged 65 or over (October
2003).
B.2 Distributional impacts of the reforms
In the tables in this appendix, we show the distributional impacts of the
following:
Working Tax Credits, The Modernisation of Britain’s Tax and Benefit System no. 10, London,
2002 (www.hm-treasury.gov.uk/mediastore/otherfiles/new_tax_credits.pdf), explains how the
new tax credits will work; Department for Work and Pensions, Pension Credit: The
Government’s Proposals, London, 2001
(www.dwp.gov.uk/publications/dwp/2001/pencred/pencred.pdf), and T. Clark, Rewarding
Saving and Alleviating Poverty? The Final Pension Credit Proposals, Briefing Note no. 22,
IFS, London, 2002 (www.ifs.org.uk/pensions/bn22.pdf), explain the details of the pension
credit; for details of all benefit rates from April 2003, see Department for Work and Pensions,
‘Pensioners and New Mothers Set to Gain from Cash Gains’, Press Release, 18 November
2002, www.dwp.gov.uk/mediacentre/pressreleases/2002/nov/cfd1811-uprt.htm.
5
1.7% was the headline rate of inflation (RPI) in September 2002, which is the rate usually
used to adjust tax allowances and thresholds.
6
Families with children on income support will not be affected by the child tax credit until
April 2004: see Chapter 4.
7
This reform only applies to new claimants, but we model the long-run impact of this change
by abolishing the increases for existing claimants.
136
Appendix B
(1)
the direct tax rises that take effect in April 2003 (i.e. the freeze in
income tax personal allowances, the rises in NI rates and the freeze in
the thresholds for NI);
(2)
all reforms that take effect in April 2003 (i.e. the tax rises in (1), plus
the increases in benefits and tax credits other than the pension credit);
(3)
all reforms that take effect in 2003–04 (i.e. the tax rises in (1), plus the
benefit and tax credit increases in (2), plus the new pension credit).
In line with usual practice, we assume that the increase in NI paid by
employers is passed on to the employee in the form of lower earnings, and that
it therefore has the same impact as a rise in employee NI.8 But in some tables,
we show separately the impact of the rise in employer NI rates and the
freezing of the secondary threshold, in case readers wish to subtract it.
Tables B.1 and B.2 show the average impact on each decile of the population,
as a proportion of disposable income and in cash terms.9 Tables B.3 and B.4
show the average impact on a number of different family types. Table B.5
shows how many families gain and lose from the reforms.
Tables B.1 and B.2 show the following:
•
The tax rises are progressive: the rich, on average, lose a greater
proportion of their income than the poor.
•
Just under half the approximately £8 billion raised by the NI and income
tax increases is offset by the extra spending on benefits and tax credits in
April 2003, rising to over 60% once the pension credit is included.
•
In the top two income deciles, the change in employer NI accounts for
around half of the average loss.
•
All but the richest tenth of families will see noticeable average gains from
the new tax credits and other benefit changes, but the impact is greater in
the lower income deciles.
•
The combined impact of a tax rise that hurts the rich more than the poor,
and new credits that benefit the poor more than the rich, is clearly
progressive. This is in keeping with the distributional impact of personal
tax and benefit changes since 1997.10 On average, the bottom five deciles
gain and the top five deciles lose, with decile 2 gaining the most.11
8
This assumption is discussed in more detail in Chapter 9.
9
Gains and losses are shown per week; the difference between full-year and part-year changes
discussed in footnote 1 is therefore not an issue. See Table B.6 for the income bands
corresponding to each decile.
10
See figure 3 in M. Brewer, T. Clark and M. Wakefield, ‘Social security in the UK under
New Labour: what did the third way mean for welfare reform?’, Fiscal Studies, 2002, vol. 23,
pp. 505–37.
11
If we do not consider the impact of raising employer NI, the bottom six income deciles gain
on average and the top four deciles lose on average.
137
Green Budget, January 2003
Table B.1. Percentage gains across the income distribution from reforms
in 2003–04
Direct tax
All changes in
All changes in
Employer NI
changes in
April 2003
2003–04
only
April 2003
Poorest
–0.08
3.43
3.94
–0.02
Decile 2
–0.14
3.24
5.62
–0.06
Decile 3
–0.43
2.65
4.08
–0.22
Decile 4
–0.73
1.32
2.52
–0.35
Decile 5
–1.04
–0.02
0.79
–0.50
Decile 6
–1.38
–0.71
–0.23
–0.67
Decile 7
–1.58
–1.11
–0.87
–0.77
Decile 8
–1.89
–1.60
–1.55
–0.94
Decile 9
–2.10
–1.99
–1.98
–1.06
Richest
–2.35
–2.35
–2.35
–1.13
–1.62
–0.84
–0.42
–0.78
Overall
Notes: The first three columns include the effects of employer NI; this is given separately in
the final column and can be subtracted if required. Income deciles are derived by dividing all
families (with and without children) into 10 equally sized groups according to income
adjusted for family size using the McClements equivalence scale. Decile 1 contains the
poorest tenth of the population, decile 2 the second poorest and so on, up to decile 10, which
contains the richest tenth.
Source: IFS tax and benefit model, TAXBEN, based on 2000–01 Family Resources Survey.
Table B.2. Weekly cash gains across the income distribution from reforms
in 2003–04 (2002 prices)
Direct tax
changes in
April 2003
Poorest
–0.06
Decile 2
–0.21
Decile 3
–0.77
Decile 4
–1.48
Decile 5
–2.41
Decile 6
–3.71
Decile 7
–4.97
Decile 8
–7.10
Decile 9
–9.78
Richest
–19.95
–5.04
Overall
Notes: As for Table B.1.
Source: As for Table B.1.
All changes in
April 2003
All changes in
2003–04
Employer NI
only
2.46
4.77
4.82
2.68
–0.05
–1.92
–3.47
–6.03
–9.27
–19.96
–2.60
2.83
8.28
7.40
5.12
1.82
–0.63
–2.74
–5.82
–9.23
–19.95
–1.29
–0.01
–0.09
–0.40
–0.69
–1.09
–1.68
–2.25
–3.26
–4.49
–8.53
–2.25
Tables B.3 and B.4 show the following:
•
On average, families with someone in work are net losers, with pensioners
and those not working gaining.
•
On average, families with children fare better than those without. For
example, single-earner couples with children will lose an average of £9.81
a week from the tax rises, but gain an average of £6.70 a week from the
new tax credits. This will leave them an average of £3.11 a week worse
off, compared with an average loss of £4.33 for single-earner couples
without children, who tend to gain less from the new tax credits.
138
Appendix B
Table B.3. Percentage gains from reforms in 2003–04, by family type
Direct tax
All changes in
All changes in
Employer NI
changes in
April 2003
2003–04
only
April 2003
Single, not working
–0.05
0.08
0.14
0.00
Single, employed
–2.25
–1.89
–1.89
–1.14
Single-parent family
–0.61
4.14
4.15
–0.33
0-earner couple w/o kids
–0.10
0.27
0.69
–0.01
0-earner couple with kids
–0.03
5.16
5.19
0.00
1-earner couple w/o kids
–1.87
–1.21
–1.18
–0.90
1-earner couple with kids
–2.06
–0.65
–0.65
–1.01
2-earner couple w/o kids
–2.32
–2.24
–2.24
–1.13
2-earner couple with kids
–2.22
–1.64
–1.64
–1.10
Single pensioner
–0.03
1.06
3.99
–0.02
Couple pensioner
–0.09
0.67
2.59
–0.05
–1.62
–0.84
–0.42
–0.78
Overall
Notes: The first three columns include the effects of employer NI; this is given separately in
the final column and can be subtracted if required. Families are classified as pensioners if
either adult is a pensioner (male: 65 or over; female: 60 or over). Some of these families also
contain children; some male recipients of the MIG can therefore be found in non-pensioner
families. See Table B.7 for the number of families of each type.
Source: IFS tax and benefit model, TAXBEN, based on 2000–01 Family Resources Survey.
Table B.4. Weekly cash gains from reforms in 2003–04, by family type
(2002 prices)
Direct tax
changes in
April 2003
Single, not working
–0.04
Single, employed
–5.29
Single-parent family
–1.46
0-earner couple w/o kids
–0.25
0-earner couple with kids
–0.07
1-earner couple w/o kids
–6.86
1-earner couple with kids
–9.81
2-earner couple w/o kids
–12.55
2-earner couple with kids
–12.53
Single pensioner
–0.06
Couple pensioner
–0.27
–5.04
Overall
Notes: As for Table B.3.
Source: As for Table B.3.
All changes in
April 2003
All changes in
2003–04
Employer NI
only
0.08
–4.45
9.91
0.70
13.90
–4.45
–3.11
–12.15
–9.26
1.88
2.12
–2.60
0.13
–4.44
9.94
1.78
13.98
–4.33
–3.11
–12.12
–9.26
7.07
8.24
–1.29
0.00
–2.41
–0.77
–0.02
0.00
–3.00
–4.33
–5.51
–5.59
–0.03
–0.15
–2.25
Table B.5 shows the following:
•
The new tax credits sharply reduce the number of lone parents who would
otherwise have suffered significant income losses as a result of the tax
increases – the number of lone parents more than £1 a week worse off falls
from 540,000 to 130,000, and 1.6 million will be net gainers. The number
of couples with children losing more than £1 a week is reduced
proportionately less by the tax credits – from 4.5 million to 3.4 million –
with 1.6 million couples with children ending up net gainers.
139
Green Budget, January 2003
Table B.5. Numbers of winners and losers, by family type
Number of families
(thousands)
Single, no children
losing >£10 per week
losing £1–£10 per week
staying within +/–£1
gaining £1–£10 per week
gaining >£10 per week
Direct tax changes
in April 2003
All changes
in April 2003
All changes
in 2003–04
725
5,335
4,294
0
0
725
5,061
4,138
233
197
725
5,055
4,132
238
203
Single with children
losing >£10 per week
losing £1–£10 per week
staying within +/–£1
gaining £1–£10 per week
gaining >£10 per week
36
507
1,239
0
0
26
103
37
1,052
565
26
103
37
1,050
567
Couple, no children
losing >£10 per week
losing £1–£10 per week
staying within +/–£1
gaining £1–£10 per week
gaining >£10 per week
2,093
2,748
1,215
0
0
2,085
2,547
997
283
145
2,085
2,543
953
296
179
Couple with children
losing >£10 per week
losing £1–£10 per week
staying within +/–£1
gaining £1–£10 per week
gaining >£10 per week
2,121
2,332
728
0
0
1,817
1,608
138
935
684
1,817
1,608
138
934
685
Pensioners
losing >£10 per week
8
7
losing £1–£10 per week
190
109
staying within +/–£1
6,622
635
gaining £1–£10 per week
0
6,055
gaining >£10 per week
0
15
Total
losing >£10 per week
4,984
4,661
losing £1–£10 per week
11,113
9,427
staying within +/–£1
14,098
5,943
gaining £1–£10 per week
0
8,559
gaining >£10 per week
0
1,608
Notes: As for Table B.3. Totals may not sum due to rounding.
Source: As for Table B.3.
•
7
109
501
4,139
2,065
4,661
9,418
5,760
6,658
3,700
Of the 1.6 million lone parents who gain overall from the changes,
1.2 million earn too little to be affected by the tax rises. Amongst the
1.6 million couples with children who gain overall from the changes,
between 0.6 million and 0.7 million earn too little to be affected by the tax
rises. This means that just over a fifth of couples with children affected by
the tax rises will gain more in tax credits than they lose in extra tax
payments.
140
Appendix B
•
Around 90% of pensioner families (6 million) will gain by at least £1 a
week from the pension credit and the overindexation of the state pension
and income tax allowance.
•
A small number of working-age families without children – mostly older
workers – gain from the introduction of the new working tax credit, with
around 380,000 gaining more than £10 a week.
•
Across the whole population, around 16 million families will lose by £1 a
week or more from the tax rises due in April 2003. Accounting for all the
changes due in 2003–04, around 14 million families will lose by £1 a week
or more, and 10 million families will gain by £1 a week or more, with
6 million families relatively unaffected. Of the 10 million families that
gain, at least 8 million (mostly pensioners) are unaffected by the tax rises;
this again shows that the changes to benefits and credits are playing a
small role in compensating families that will be paying more tax.
Table B.6. Income bands for each decile for different family types (net
household income in £ per year, 2002 prices, after direct taxes and
benefits)
Single, no children
Couple, no children
Couple, two children
Poorest
0–5,400
0–8,800
0–12,600
Decile 2
5,400–6,500
8,800–10,700
12,600–15,300
Decile 3
6,500–7,600
10,700–12,500
15,300–17,900
Decile 4
7,600–8,800
12,500–14,400
17,900–20,600
Decile 5
8,800–10,100
14,400–16,600
20,600–23,700
Decile 6
10,100–11,700
16,600–19,100
23,700–27,300
Decile 7
11,700–13,500
19,100–22,200
27,300–31,800
Decile 8
13,500–16,100
22,200–26,400
31,800–37,800
Decile 9
16,100–20,400
26,400–33,400
37,800–47,800
Richest
20,400–
33,400–
47,800–
Note: There are 3,020,000 families in each decile.
Source: IFS tax and benefit model, TAXBEN, based on 2000–01 Family Resources Survey.
Table B.7. Estimated number of families in Great Britain, by family type
Number of families (millions)
Single, not working
3.3
Single, employed
7.1
Single-parent family (working or not working)
1.8
0-earner couple w/o kids
0.9
0-earner couple with kids
0.4
1-earner couple w/o kids
1.6
1-earner couple with kids
1.6
2-earner couple w/o kids
3.5
2-earner couple with kids
3.2
Single pensioner
4.2
Couple pensioner
2.7
Total
30.2
Source: IFS tax and benefit model, TAXBEN, based on 2000–01 Family Resources Survey.
141
Appendix C: Budgets since 1979
This appendix summarises the main tax measures introduced in each Budget since
1979. Statutory indexation of thresholds and limits is not included.
1979 Budget, Geoffrey Howe
Income tax
Basic rate cut from 33% to 30%.
Top rate cut from 83% to 60% on earned income and from 98% to 75% on
unearned income.
VAT
Two-tier rates of 8% and 12.5% replaced by single 15% rate.
Excise duties
Alcohol and tobacco duties reduced; petrol duty increased.
Company taxes
Petroleum revenue tax rate increased from 45% to 60%.
1980 Budget, Geoffrey Howe
Income tax
Reduced rate of 25% abolished.
National Insurance
Employee rate increased from 6.5% to 6.75% (contracted in).
Employer rate increased from 10% to 10.2% (contracted in).
Capital taxes
Stamp duty threshold on property increased from £15,000 to £20,000.
Capital transfer tax threshold doubled from £25,000 to £50,000.
Company taxes
Petroleum revenue tax rate increased from 60% to 70%.
1981 Budget, Geoffrey Howe
Income tax
Personal allowances frozen in cash terms, implying a cut in real terms.
National Insurance
Employee rate increased from 6.75% to 7.75% (contracted in).
Excise duties
Sharp increases (beer and petrol up 24%, cigarettes up 16%).
1982 Budget, Geoffrey Howe
Income tax
Personal allowances increased in real terms.
National Insurance
Employee rate increased from 7.75% to 8.75% (contracted in).
Employer National Insurance surcharge reduced from 3.5% to 2%, and to
1.5% from April 1983.
Capital taxes
Indexation provisions introduced for capital gains tax.
Stamp duty threshold on property increased from £20,000 to £25,000.
Company taxes
Petroleum revenue tax rate increased from 70% to 75%.
1983 Budget, Geoffrey Howe
Income tax
Personal allowances increased in real terms.
Mortgage interest relief ceiling raised from £25,000 to £30,000.
National Insurance
Employee rate increased from 8.75% to 9% (contracted in).
Employer National Insurance surcharge cut from 1.5% to 1%.
Company taxes
Licence royalties abolished for all new oilfields.
1984 Budget, Nigel Lawson
Income tax
Personal allowances increased in real terms.
Investment income surcharge abolished.
Relief on life assurance premiums abolished for new policies.
National Insurance
Employer National Insurance surcharge abolished.
Excise duties
Duty on wine cut sharply; increases on beer and cigarettes.
Capital taxes
Stamp duty threshold on property increased from £25,000 to £30,000.
Highest rate of stamp duty reduced from 2% to 1%.
Top rate of capital transfer tax cut from 75% to 60%.
Company taxes
Corporation tax rate to be reduced from 52% in 1982–83 to 50% in
1983–84, 45% in 1984–85, 40% in 1985–86 and 35% in 1986–87.
Stock relief abolished.
First-year allowances to be phased out and replaced by 25% writing-down
allowances.
142
Appendix C
1985 Budget, Nigel Lawson
Income tax
Personal allowances increased in real terms.
National Insurance
Employee and employer contributions restructured, with reduced rates for
lower earners.
Upper ceiling on employer contributions abolished.
Company taxes
Development land tax abolished.
1986 Budget, Nigel Lawson
Income tax
Basic rate reduced from 30% to 29%.
Announcement of the introduction of tax relief for profit-related pay (PRP)
schemes in 1987.
Tax relief for Personal Equity Plans (PEPs) introduced.
Capital taxes
Capital transfer tax replaced with inheritance tax.
Stamp duty for shares reduced from 1% to 0.5%.
1987 Budget, Nigel Lawson
Income tax
Basic rate reduced from 29% to 27%.
Excise duties
Duties held constant in cash terms, implying a real cut.
Capital taxes
Inheritance tax threshold increased from £71,000 to £90,000.
Number of inheritance tax rates cut from seven to four.
New arrangements to encourage personal pensions.
1988 Budget, Nigel Lawson
Income tax
Personal allowances increased in real terms.
Basic rate reduced from 27% to 25%.
All rates above 40% abolished.
Announcement of separate taxation of husband and wife from 1990.
Company car scale charges doubled.
Capital taxes
Capital gains accruing before 1982 written off for capital gains tax purposes.
Capital gains tax rates changed to equal marginal income tax rates.
Inheritance tax threshold increased from £90,000 to £110,000.
Inheritance tax rates reduced to a single rate of 40%.
1989 Budget, Nigel Lawson
Income tax
Limit for higher age relief reduced to 75.
Age allowance taper reduced to 50%.
Pensioner ‘earnings rule’ abolished.
PEPs extended.
National Insurance
Employee 5% and 7% bands abolished.
Lower 2% rate for employees introduced on earnings below lower earnings
limit.
Excise duties
Petrol duties adjusted to favour unleaded fuel.
1990 Budget, John Major
Income tax
Basic-rate limit frozen.
Employer-provided work-place nurseries exempted from tax.
Introduction of Tax-Exempt Special Savings Accounts (TESSAs).
Abolition of composite rate of tax announced.
Capital taxes
Plans for abolition of stamp duty on shares announced.
Company taxes
Corporation tax rate cut from 35% to 34%.
1991 Budget, Norman Lamont
Income tax
Married couple’s allowance frozen.
Mortgage interest relief restricted to the basic rate of tax.
PEPs extended.
Company car scale charges raised by 20%.
National Insurance
Employer contributions to be charged on company cars and free fuel from
1992–93.
VAT
Standard rate of VAT raised from 15% to 17.5%.
Company taxes
Corporation tax rate cut from 34% to 33%.
143
Green Budget, January 2003
Local taxes
Community charge bills subsidised by £140 per adult.
1992 Budget, Norman Lamont
Income tax
Reduced rate of 20% introduced on first £2,000 of taxable income.
Married couple’s allowance frozen.
Basic-rate limit frozen.
PEPs limit on investment and unit trusts raised from £3,000 to the overall
limit, £6,000.
Excise duties
Further widening in leaded–unleaded petrol duty differential.
Car tax halved from 10% to 5% and abolished from November 1992.
1993 Spring Budget, Norman Lamont
Income tax
20% band widened to £3,000 by April 1994.
Personal allowances and basic-rate limit frozen.
Married couple’s allowance and mortgage interest relief restricted to 20%
from April 1994.
National Insurance
Contribution rates for employees and self-employed up 1 percentage point
from April 1994.
VAT
Extended to domestic fuel at 8% from April 1994 and at 17.5% from April
1995.
Excise duties
Duties increased above inflation, except spirits (frozen).
Announced commitment to increase duties on road fuel by at least 3% p.a.
in real terms.
Capital taxes
Stamp duty threshold doubled to £60,000.
Company taxes
Advance corporation tax (ACT) rate reduced to 22.5% from April 1993 and
to 20% from April 1994.
Dividend ‘tax credit’ down to 20%.
Basic rate of tax on dividends reduced to 20%.
Local taxes
Community charge abolished, council tax introduced.
1993 Autumn Budget, Kenneth Clarke
Income tax
Personal allowances and basic-rate limit frozen.
Married couple’s allowance and mortgage interest relief restricted to 15%
from April 1995.
National Insurance
Main rate for employer contributions reduced by 0.2 of a percentage point to
10.2%.
Lower rates of employer contributions reduced by 1 percentage point.
Excise duties
No increase on spirits and beer.
Most other duties increased above indexation.
Commitment to raise tobacco duties by at least 3% p.a. in real terms.
Commitment to raise road fuel duties by at least 3% p.a. in real terms
increased to 5% p.a. in real terms.
Insurance premium tax and air passenger duty introduced.
1994 Budget, Kenneth Clarke
Income tax
All age-related personal allowances increased above inflation.
VAT
Abandonment of second stage of VAT on domestic fuel – rate to stay at 8%.
Excise duties
Alcohol duties raised by an average of 4%.
Tobacco duties increased by more than inflation.
Duties on road fuel increased above inflation; diesel duties brought in line
with duties on unleaded petrol.
Other
Landfill tax planned for 1996 and businesses to be compensated through
lower employer National Insurance contributions.
1995 Budget, Kenneth Clarke
Income tax
Basic rate of income tax reduced from 25% to 24%.
Personal allowances increased above inflation.
Lower-rate band and basic-rate limit increased by more than indexation.
Tax on savings income cut from 25% to 20% for basic-rate taxpayers.
144
Appendix C
National Insurance
Excise duties
Capital taxes
Company taxes
Other taxes
Tax relief on Class 4 National Insurance contributions withdrawn.
Main rate for employers cut from 10.2% to 10% from April 1997.
Rate of Class 4 contributions reduced from 7.3% to 6%.
Beer, wine and most cider duties frozen; spirits cut by 4%.
Most tobacco duties up by 3% in real terms.
Petrol and diesel tax (duty and VAT) raised by 3.5p per litre, or 5% real
increase.
Inheritance tax threshold raised to £200,000, £40,000 more than indexation.
Small companies’ rate cut from 25% to 24%.
Landfill tax introduced at two rates, of £2 and £7 per tonne.
1996 Budget, Kenneth Clarke
Income tax
Personal allowances increased by more than inflation.
Basic-rate limit and married couple’s allowance indexed.
Basic rate cut to 23%.
Tax relief for profit-related pay phased out from 1998–99.
Excise duties
Beer, wine and cider duties frozen; duty on spirits cut by 4%.
Air passenger duty doubled, insurance premium tax up to 4%.
Tobacco up by 5% in real terms, hand-rolling tobacco indexed.
Petrol and diesel up by 5% in real terms.
Company taxes
Small companies’ rate cut to 23%.
Capital allowances cut for long-lived assets.
Local taxes
Transitional relief for small companies extended.
1997 Summer Budget, Gordon Brown
Income tax
Mortgage interest relief cut to 10% from April 1998.
VAT
Rate on domestic fuel cut from 8% to 5%.
Excise duties
Road fuel duties commitment raised from 5% p.a. to 6% p.a. real increase.
Tobacco duty commitment raised from 3% p.a. to 5% p.a. real increase.
Capital taxes
Graduated stamp duty introduced: 1% for properties between £60,000 and
£250,000; 1.5% between £250,000 and £500,000; 2% over £500,000.
Company taxes
Windfall tax on privatised utilities.
Main corporation tax rate cut from 33% to 31% from April 1997.
Small companies’ rate cut from 23% to 21% from April 1997.
Dividend tax credits for pension funds and other companies abolished
immediately, for all others from April 1999.
1998 Spring Budget, Gordon Brown
Income tax
Working families’ tax credit from October 1999.
Married couple’s allowance restricted to 10% from April 1999.
Individual Savings Accounts (ISAs) from April 1999.
Tax on company cars increased.
National Insurance
‘Entry fee’ abolished for employees from April 1999.
Excise duties
Differential widened between diesel and unleaded petrol.
Capital taxes
Personal capital gains tax reformed: indexation abolished and taper
introduced.
Stamp duty raised to 2% on properties between £250,000 and £500,000, 3%
on properties over £500,000.
Company taxes
ACT abolished from April 1999 and quarterly payments system introduced.
Main rate cut to 30%, small companies’ rate to 20% from April 1999.
1999 Budget, Gordon Brown
Income tax
Basic rate cut from 23% to 22% from April 2000.
New 10% starting rate from April 1999; 20% rate abolished.
Married couple’s allowance abolished from 2000 for under-65s.
Children’s tax credit announced from April 2001.
Mortgage interest relief abolished from April 2000.
High mileage discounts for company cars reduced.
145
Green Budget, January 2003
National Insurance
Capital taxes
Company taxes
Starting point for payment of employee National Insurance contributions
aligned with income tax by April 2001.
Upper earnings limit raised above inflation in both April 2000 and April
2001.
Self-employed structure reformed from April 2000.
Employer contributions on all benefits in kind.
Employer rate cut by 0.5 of a percentage point from April 2001.
Stamp duty raised to 2.5% on properties between £250,000 and £500,000,
3.5% on properties over £500,000.
Climate change levy from 2001–02.
2000 Budget, Gordon Brown
Income tax
Working families’ tax credit and child premiums in children’s tax credit
increased.
National Insurance
Employer rate to be cut by 0.3 of a percentage point from April 2001,
instead of 0.5 of a percentage point, to reflect reduction in climate change
levy.
Further cut in employer rate by 0.1 of a percentage point from April 2002, to
balance introduction of aggregates levy.
Excise duties
Road fuel duty frozen in real terms.
Cigarettes increased by 5% in real terms.
Capital taxes
Stamp duty raised to 3% on properties between £250,000 and £500,000, 4%
on properties over £500,000.
Company taxes
Climate change levy cut by £0.7 billion from introduction in April 2001.
Aggregates levy introduced from April 2002.
2001 Budget, Gordon Brown
Income tax
Working families’ tax credit and child premiums in children’s tax credit
increased.
Overindexation of starting-rate band.
ISA limit extended to £7,000 p.a. until April 2006.
Excise duties
Duties for ultra-low sulphur petrol cut by 2p and for ultra-low sulphur diesel
cut by 3p.
Tobacco duties increased with inflation; alcohol duties frozen.
Company taxes
Abolition of withholding tax on intra-UK corporate interest.
2002 Budget, Gordon Brown
Income tax
Child tax credit introduced to replace various income-related payments for
children. Working tax credit introduced for both families with and families
without children; working families’ tax credit abolished.
Personal allowances for those aged under 65 to be frozen in cash terms in
April 2003.
National Insurance
Uncapped 1 percentage point increase in employee, employer and selfemployed rates from April 2003.
Primary and secondary thresholds and lower profits limit to be frozen in
cash terms in April 2003.
Excise duties
Fuel duties frozen in cash terms.
Company taxes
Small companies’ rate cut from 20% to 19%.
Starting rate of corporation tax reduced from 10% to 0%.
Research and development tax credit introduced for larger companies at
25% rate.
Reform to North Sea taxation.
146
Appendix D: Headline tax and
benefit rates and thresholds
2002–03 level
2003–04 levela
Income tax
Personal allowance: under age 65
aged 65–74
aged 75 and over
Married couple’s allowance, restricted to 10%:
aged 65 or over on 6 April 2000
aged 75 or over
Lower rate
Basic rate
Higher rate
Lower-rate limit
Basic-rate limit
Pension earnings cap
Tax rates on interest income
Tax rates on dividend income
Children’s tax creditb
Children’s tax credit for first year of a child’s lifeb
£4,615 p.a.
£6,100 p.a.
£6,370 p.a.
£4,615 p.a.
£6,610 p.a.
£6,720 p.a.
£5,465 p.a.
£5,535 p.a.
10%
22%
40%
£1,920 p.a.
£29,900 p.a.
£97,200 p.a.
10%, 20%, 40%
10%, 32.5%
£5,290 p.a.
£10,490 p.a.
£5,565 p.a.
£5,635 p.a.
10%
22%
40%
£1,960 p.a.
£30,500 p.a.
£98,900 p.a.
10%, 20%, 40%
10%, 32.5%
Replaced by
child tax credit
National Insurance
Lower earnings limit (LEL)
Upper earnings limit (UEL)
Earnings threshold (employee and employer)
Class 1 contracted-in rate: employee – below UEL
– above UEL
employer – below UEL
– above UEL
Class 1 contracted-out rate: employee – below UEL
(salary-related schemes)
– above UEL
employer – below UEL
– above UEL
£75 p.w.
£585 p.w.
£89 p.w.
10%
zero
11.8%
11.8%
8.4%
zero
8.3%
11.8%
£77 p.w.
£595 p.w.
£89 p.w.
11%
1%
12.8%
12.8%
9.4%
1%
9.3%
12.8%
zero
19%
30%
zero
19%
30%
£7,700 p.a.
£3,850 p.a.
24%–40%
12%–20%
10%–40%
5%–20%
£7,900 p.a.
£3,950 p.a.
24%–40%
12%–20%
10%–40%
5%–20%
£250,000
40%
£255,000
40%
17.5%
5%
17.5%
5%
Corporation tax
Rates: lower rate
small companies’ rate
standard rate
Capital gains tax
Annual exemption limit: individuals
trusts
Non-business assets: top-rate taxpayers
basic-rate taxpayers
Business assets: top-rate taxpayers
basic-rate taxpayers
Inheritance tax
Threshold
Rate for transfer at or near death
Value added tax
Standard rate
Reduced rate
Continues
147
Green Budget, January 2003
Continued
2002–03 level
2003–04 levela
Excise duties
Beer (pint)
Wine (75cl bottle)
Spirits (70cl bottle)
20 cigarettes: specific duty
ad valorem (22% of retail price)
Ultra-low sulphur petrol (litre)
Ultra-low sulphur diesel (litre)
26p
116p
548p
188p
92p
46p
46p
27p
118p
556p
191p
93p
47p
47p
Air passenger duty
Destinations within the EU: economy
club/first class
Destinations outside the EU: economy
club/first class
£5
£10
£20
£40
£5
£10
£20
£40
Betting and gaming duty
Gross profits tax
Spread betting rate: financial bets
other bets
15%
3%
10%
15%
3%
10%
5%
17.5%
5%
17.5%
£60,000 p.a.
0%
1%
3%
4%
0.5%
£60,000 p.a.
0%
1%
3%
4%
0.5%
Insurance premium tax
Standard rate
Higher rate (for insurance sold accompanying certain
goods and services)
Stamp duty
Land and buildings: threshold
rate: up to £60,000
£60,000–£250,000
£250,000–£500,000
above £500,000
Stocks and shares: rate
Vehicle excise duty
Graduated system for new cars from 1 March 2001
Standard rate
Small-car rate (engines up to 1,549cc)
Heavy goods vehicles (varies according to vehicle type
and weight)
Landfill levy
Standard rate
Low rate (inactive waste only)
£60–£160 p.a.
£160 p.a.
£105 p.a.
£160–£1,850 p.a.
£13 per tonne
£2 per tonne
Local taxes
Average rate band D council tax: England
£14 per tonne
£2 per tonne
£976
Income support / income-based jobseeker’s allowance
Single (aged 25 or over)
Couple (both aged 18 or over)
Family premium
Child allowance: aged under 16
aged 16–18
Minimum income guarantee for those aged 60 or over:
single
couple
Pension credit taper (from October 2003)
Winter fuel payment for those aged 60 or over
148
£53.95 p.w.
£84.65 p.w.
£14.75 p.w.
£37.00 p.w.
£37.80 p.w.
£54.65 p.w.
£85.75 p.w.
£15.75 p.w.
£38.50 p.w.
£38.50 p.w.
£98.15 p.w.
£149.80 p.w.
n/a
£200
£102.10 p.w.
£155.80 p.w.
40%
£200
Continues
Appendix D
Continued
2002–03 level
2003–04 levela
Child benefit
First child
Other children
£15.75 p.w.
£10.55 p.w.
£16.05 p.w.
£10.75 p.w.
Basic state pension
Single
Couple
£75.50 p.w.
£120.70 p.w.
£77.45 p.w.
£123.80 p.w.
Working families’ tax credit
Basic (adult) credit
30-hour credit
Child credit: aged under 16
aged 16–18
Disabled child credit
Applicable amount (i.e. threshold or earnings disregard)
Childcare tax credit:
maximum eligible cost for one child
maximum eligible cost for two or more children
proportion of eligible costs covered
Child tax credit
Family element
Family element for first year of a child’s life
Child element
Disabled child element
Working tax credit
Basic element
Couples and lone-parent element
30-hour element
Disabled worker element
Childcare element:
maximum eligible cost for one child
maximum eligible cost for two or more children
proportion of eligible costs covered
Features common to child and working tax credits
First income threshold
First income threshold if entitled to child tax credit only
First withdrawal rate
Second income threshold
Second withdrawal rate
£62.50 p.w.
£11.65 p.w.
£26.45 p.w.
£27.20 p.w.
£35.50 p.w.
£94.50 p.w.
Replaced by
child tax credit
and working tax
credit
£135.00 p.w.
£200.00 p.w.
70%
n/a
n/a
n/a
n/a
£545 p.a.
£1,090 p.a.
£1,445 p.a.
£2,155 p.a.
n/a
n/a
n/a
n/a
£1,525 p.a.
£1,500 p.a.
£620 p.a.
£2,040 p.a.
n/a
n/a
n/a
£135.00 p.w.
£200.00 p.w.
70%
n/a
n/a
n/a
n/a
n/a
£5,060 p.a.
£13,230 p.a.
37%
£50,000 p.a.
1 in 15
Maternity benefits
Sure Start maternity grant
£500
£500
Statutory maternity pay: weeks 1–6
90% earnings
90% earnings
weeks 7–18
£75.00 p.w.
£100.00 p.w.
weeks 19–26
zero
£100.00 p.w.
Maternity allowance: weeks 1–18
£75.00 p.w.
£100.00 p.w.
weeks 19–26
zero
£100.00 p.w.
a
2003–04 figures assume no discretionary changes apart from pre-announced measures and statutory
indexation where appropriate.
b
Allowance is available at a flat rate of 10% and is tapered away from families that include a higherrate taxpayer.
Sources: Various HM Treasury, Inland Revenue and HM Customs and Excise Press Releases, April
2002 and November 2002; HM Treasury, Tax Ready Reckoner and Tax Reliefs, London, 2002
(www.hm-treasury.gov.uk/media//DB8B5/adtrr02.pdf); www.inlandrevenue.gov.uk/rates;
www.hmce.gov.uk/business/othertaxes/othertaxes.htm; www.dwp.gov.uk/lifeevent/benefits/index.htm;
www.local.dtlr.gov.uk/finance/ctax/ctax023.htm; Tolley, Tolley’s Tax Data 2002–03, London, 2002.
149
Appendix E: Tax revenues ready reckoner
Table E.1. Direct effects of illustrative changes in taxation to take effect April 2003
Cost/yield
(non-indexed base)
2003–04
(£m)
Income tax
Rates
Change starting rate by 1pa
Change basic rate by 1pb
Change higher rate by 1p
Change basic rate in Scotland by 1p
560
3,250
1,120
260
Allowances
Change personal allowance by £100
620
Starting-rate limit
Change starting-rate limit by £100
320
Basic-rate limit
Change basic-rate limit by 1%
Change basic-rate limit by 10%:
increase (cost)
decrease (yield)
1,700
2,100
Allowances and limits
Change all main allowances, starting- and basic-rate limits:
increase/decrease by 1%
increase by 10% (cost)
decrease by 10% (yield)
590
5,600
6,300
190
National Insurance
Rates
Change Class 1 employee rate between entry threshold and upper earnings limit
by 1 percentage point
Change Class 1 employee rate above upper earnings limit by 1 percentage point
Change Class 1 employer rate by 1 percentage point
Change Class 2 (self-employed) rate by £1 a week
Change Class 4 (self-employed) rate by 1 percentage point
Allowances
Change employee entry threshold by £2 per week
Change employer entry threshold by £2 per week
Change upper earnings limit by £10 per week
a
Including savings income taxable at the starting rate, but excluding dividend income.
b
Covering savings income (but not the starting rate), and excluding dividends.
3,150
655
3,950
130
290
245
290
125
Continues
150
Appendix E
Continued
Cost/yield
(non-indexed base)
2003–04
(£m)
Corporation tax
Change main rate by 1 percentage point
Change small companies’ rate by 1 percentage point
1,150
240
Capital gains tax
Increase annual exempt amount by £500 for individuals and £250 for trustees
10
Inheritance tax
Change rate by 1 percentage point
Increase threshold by £5,000
75
65
Excise dutiesc
Beer up 0.3p a pint
Wine up 1.4p a bottle (75cl)
Spirits up 6.4p a bottle (70cl)
Cigarettes up 3.6p a packet (20 king-size)
Petrol up 0.5p a litre
Diesel (ultra-low sulphur) up 0.5p a litre
Change insurance premium tax (both standard and higher rates)
by 1 percentage point
35
15
5
65
120
110
315
VAT
Change both standard and reduced rates by 1 percentage point
3,840
2002–03
Extend VAT to:
9,350
food
3,400
construction of new homes
1,750
domestic and international (UK portion) passenger transport
1,450
books, newspapers and magazines
950
water and sewerage services
800
children’s clothing
800
drugs and supplies on prescription
c
Figures are calculated given the price and tax charged on a typical item. All changes are assumed to be
implemented in April 2003.
Note: The revenue effect is computed for changes to the 2003–04 tax system and relates to the full-year
effect, except for changes to the coverage of VAT which refer to 2002–03
Source: HM Treasury, Tax Ready Reckoner and Tax Reliefs, London, November 2002 (www.hmtreasury.gov.uk/mediastore/otherfiles/adtrr02.pdf).
VAT coverage
151