The real impact of Budget 2010
Tim Horton (Fabian Society) and Howard Reed (Landman Economics)
27 June 2010
A briefing paper from a research project for the TUC and UNISON, Standing up for public services.
Foreword
This briefing paper is part of an ongoing research project for the TUC and UNISON, looking at the distribution of public spending in the UK.
The calculations in this paper are based on a new model of how public spending is allocated across households, created for the project. It allows us, for the first time in the UK, to show how all public spending is distributed across households when you take into account how often they use services. In turn, we can use this model to analyse the distributional impact of spending cuts on households.
The final report of the project, Standing up for public services,including more details of the model, will be published later in summer 2010.
About the authors
Tim Horton is Research Director at the Fabian Society, Britain’s leading left of centre think tank and political society. His areas of research expertise include social policy, economic and fiscal policy, political parties and democratic reform, and political philosophy. Prior to working at the Fabian Society, Tim was a Special Adviser at the Department of Trade and Industry, and before that a policy analyst in HM Treasury.
Howard Reed is Director of Landman Economics, a consultancy that specialises in economic research from a progressive perspective. His areas of research expertise include microeconometrics, tax and benefits and the labour market. Prior to forming Landman Economics, Howard was Chief Economist at the Institute for Public Policy Research.
Don’t forget the spending cuts! The real impact of Budget 2010
Executive summary
This report reveals the true impact on households of last week’s budget.
To date, assessments of budget’s impact, and how fair or unfair it might be, have centred on the impact of the tax and benefit changes announced, while ignoring the impact of spending cuts.
Using a new model of how public spending is allocated across households in the UK, this report estimates the distributional impact of the spending cuts announced in the budget.
Then, by combining this with government data on the impact of the budget’s tax and benefit changes, the report calculates the budget’s overall impact.
In summary, the report finds that:
- Excluding benefit cuts, the budget committed to £34 billion of spending cuts by 2012-13.
- The impact of these cuts will be deeply regressive. All households are hit considerably, but the poorest households are hit the hardest.
- Assuming these cuts fall evenly across non-ringfenced departments, the average annual cut in public spending on the poorest tenth of households is £1,344, equivalent to 20.5% of their household income, whereas the average annual cut in public spending on the richest tenth of households is £1,135, equivalent to just 1.6% of their household income
- When the impact of these spending cuts is combined with the Government’s own analysis of the impact of the budget’s tax and benefit changes – in order to generate a picture of the budget’s overall impact – the result is, once again, deeply regressive. In particular, the magnitude of the impact of spending cuts on households dwarfs the impact of the tax and benefit changes.
- Overall, the combined average annual loss in income and services for the poorest tenth of households is £1,514, equivalent to 21.7% of their household income. For the richest tenth of households, the annual loss in income and services is £2,685, equivalent to just 3.6% of their household income.
These results are illustrated in the graph below.
Despite committing to severe public spending cuts, the budget also announced an income tax cut, costing £3.7 billion, which was billed as helping low-income households in particular.
This report also looks at the net impact of this tax cut. Assuming that, in the absence of the tax cut, this £3.7 billion would have otherwise been spent on public services, we look at who really gains from the tax cut when the impact of the corresponding £3.7 billion spending cut is also taken into account.
In summary, we find that:
- The income tax cut provides much smaller proportionate gains for lower-income households than for middle- and upper-middle-income households. By contrast, lower-income households lose most heavily from the corresponding spending cut, with smaller proportionate losses for middle- and upper-middle-income households.
- The result is that many low-income households are ‘net losers’ from this income tax cut, when the corresponding reduction in public spending is taken into account. Many low-income households might well have been better off if the revenue had instead been put into protecting public services.
1) Assessing the impact of the budget: what’s fair?
The Government have previously claimed that the impact of the budget is “fair” and “progressive” on the basis of an analysis of the distributional impact of the budget’s tax and benefit changes – that is, how these changes affect households at different points on the income spectrum.
In particular, ministers have made a great play of one graph in the budget report (Chart A2, p.67), showing the distributional impact of tax and benefit changes by 2012-13. This graph (replicated by the red bars in the Graph 1 below) shows that, by 2012-13, the poorest tenth of households lose an amount equivalent to 1% of their income from the budget’s tax and benefit changes, whereas the richest tenth of households lose an amount equivalent to 2% of their income. Ministers claim this shows that the budget meets the fairness test of hitting those on higher incomes more than those on lower incomes.
But this completely ignores the impact of cuts in public spending on households. What really counts for fairness is not how families are affected by tax and benefit changes in isolation, but how they are affected by the whole package – spending cuts included.
It’s not just that coalition ministers don’t want to talk about spending cuts. The trouble is, while organisations like the Treasury and the Institute for Fiscal Studies (IFS) have long had the capability to calculate the distributional impact of tax and benefit changes, it’s much harder to do this for spending on public services, not least because it’s difficult to work out who uses different services and how much they use them.
As Robert Chote, the Director of the IFS put it in the IFS’ post-budget analysis, “perhaps the most important omission in any distributional analysis of this sort is the impact of the looming cuts to public services, which are likely to hit poorer households significantly harder than richer households.”
2) Analysing the distributional impact of public spending
To fill this gap, we have created a new model for analysing how public spending is allocated across households. It shows, for the first time in the UK, how all public spending on services is distributed amongst households when you take into account how often households use services.
The model is based on extensive analysis of household-level surveys, conducted by the Office for National Statistics, which contain information on public service usage (for example, the General Household Survey and the Expenditure and Food Survey). We have then reconciled this information with government spending data (using an Expenditure-on-Services budgeting approach via HM Treasury’s Public Expenditure Statistical Analysis framework).
Where possible, our model allocates spending to households in proportion to their actual service usage. Our model also takes account of various policy-related constraints on how particular types of spending are distributed, such as means testing. And spending in some areas where services are enjoyed ‘collectively’ or where there is no basis for allocating different amounts to different households (such as with spending on national defence) is allocated on a flat-rate basis.
In this report, we use this model of how public spending is allocated across households to show the distributional impact of spending cuts on households. To do this, we make the further assumption that spending cuts on any particular service affect households in proportion to the amount that they actually use that service.
3) The impact of spending cuts
Excluding benefit cuts (the impact of which is already included in the Government’s own analysis of tax and benefit changes used here), the budget announced a further £34 billion of spending cuts by 2012-2013. Budget Table 1.1 (page 15) shows £42 billion of spending cuts by 2012-13. We discount £3 billion of this, which are reductions in debt interest payments. (A £1 billion reduction in debt interest payments results from discretionary measures announced in the emergency budget (see Table 1, page 4), and a £2 billion reduction results from policy measures inherited from the Labour government (Footnote 2, Table 1.1, page 15). We then subtract a further £5 billion of welfare cuts, all of which results from discretionary measures announced in the emergency budget (Table 1.1, page 4). Benefit cuts are discounted in deriving this £34 billion total for spending cuts since the impact of (most of) these welfare cuts is already included in the Government’s own analysis of tax and benefit changes used in this report.
Our analysis here assumes that spending on health and international development will be protected from these cuts (ringfenced), as the Government have announced.
We therefore model the impact of these £34 billion cuts across all areas of non-benefit spending excluding health and international development. Beyond this, we assume that these cuts fall evenly across all non-ringfenced departments.
The blue bars in Graph 1 below show the distributional impact of these spending cuts by equivalised household income decile.
As can be seen from the graph, the impact of these cuts will be deeply regressive. All households are hit considerably, but the poorest households are hit the hardest. The average annual cut in public spending on the poorest tenth of households is £1,344, equivalent to 20.5% of their household income, whereas the average annual cut in public spending on the richest tenth of households is £1,135, equivalent to just 1.6% of their household income.
One important reason for this is that alot of public spending is ‘pro-poor’, with poorer households receiving a greater value of services to meet their extra welfare needs. Because of this, cuts in public spending on major areas of welfare (such as education or social housing) will tend to hit the poorest hardest.
Another important reason for this regressive impact is that, for a given value of services lost, the impact will be larger relative to household income for poorer households than for richer households. Imagine, for example, households had to replace the value of lost services by increasing their own expenditure by a fixed amount (for example, if public financing of roads through progressive taxation was replaced by private financing of roads through flat-rate tolls and user-charges). This would clearly hit lower-income households proportionately much harder than higher-income households.
Note that our calculations assume these £34 billion of spending cuts will fall evenly and proportionately across non-ringfenced areas of spending. This is a simplifying assumption, since it may be the case that some areas will suffer proportionately more severe cuts than others. But it is also a necessary assumption at the moment, since we will not know how these cuts are to be implemented until the Comprehensive Spending Review this autumn.
However, we did test some alternative scenarios, in which some departments received higher percentage cuts to their budgets than others. (For example, the government have suggested that defence spending may do better than other areas – we looked at this scenario). We found that, while different patterns of cuts of course changed the underlying figures to some extent, the overall regressive gradient of the cuts across the income distribution remained quite similar.
Because of this, and because until the autumn spending review we have no basis for allocating different levels of cuts to different departments, the results we show here are for a ‘baseline scenario’ in which all non-ringfenced departments receive the same proportional cut in their budgets. We shall update the picture in the autumn when we have a more detailed picture of where the cuts will fall.
4) The overall impact of the budget
We then combine this analysis of the distributional impact of spending cuts with data (already published by the Government in the budget report) on the impact of the budget’s tax and benefit changes. This gives a true assessment of the budget’s overall distributional impact on households – and is shown by the green line in Graph 1 below.
(Distributional impact of all tax and spending changes by 2012-13)
As can be seen from the graph, when you add in the impact of the spending cuts, the picture changes completely from when consideration is limited to the impact of tax and benefit changes alone. In particular, the magnitude of the impact of spending cuts on households dwarfs the impact of the tax and benefit changes.
Once again, the overall impact is deeply regressive, with households hit proportionately less the further you move up the income spectrum.
Overall, the combined average annual loss in income and services for the poorest tenth of households is £1,514, equivalent to 21.7% of their household income. For the richest tenth of households, the average annual loss in income and services is £2,685, equivalent to just 3.6% of their household income.
This analysis therefore undermines the Government’s claim that the impact of the budget was ‘fair’ and ‘progressive’. When you add in the impact of spending cuts, the poorest households are hit hardest, and that is deeply unfair.
Additionally, while it is likely there would have been significant spending cuts whoever had been in government, this analysis of the impact of spending cuts raises real questions about the coalition government’s decision to relymuch more heavily on spending cuts for reducing the deficit than other parties had planned to.
Far from being ‘unavoidable’, this was a discretionary decision – and one that has clearly hit low-income households much harder than they otherwise would have been.
5) The net impact of the budget tax cut: did it really help those on low incomes?
Despite committing to severe public spending cuts, the budget also announced an income tax cut, costing £3.7 billion. The tax cut in question was a £1,000 increase in the personal allowance (from £6,475 to £7,475).
This was billed as helping low-income households in particular (and formed another part of the Government’s claim that the budget was ‘fair’ and ‘progressive’).
However, this assessment completely ignores the fact that the £3.7 billion cost of this tax cut has come in the form of reduced public spending. And it is wrong simply to look at the gains from a tax cut without also looking at the losses arising from the resulting fall in public spending.
Being able to analyse the distributional impact of spending cuts enables us to calculate the net impact of a tax cut, balancing any gains to households from the tax cut against the losses incurred from the cut in public spending.
Here, we use our model to analyse the net impact of the budget’s income tax cut. Assuming that, in the absence of the tax cut, this £3.7 billion would have otherwise been spent on public services, we can look at who really gains from the tax cut when the impact of this corresponding £3.7 billion spending cut is also taken into account.
Because until the autumn spending review we have no basis for allocating different levels of cuts to different departments, once again here we assume that the £3.7 billion reduction in public spending to pay for the income tax cut falls evenly and proportionately across all service areas (except for health and international development) in working out what the overall impact on households would be. Again, then, this is a ‘baseline scenario’ that will be updated in the autumn when we have a more detailed picture of where the cuts will fall.
The graph below illustrates the distributional impact of the £3.7 billion tax cut in isolation (red bars), the impact of the corresponding £3.7 billion spending reduction (blue bars), and finally the net impact of the measure (green line) – combining the two effects.
As can be seen, the income tax cut provides much smaller proportionate gains for lower-income households than for middle- and upper-middle-income households. By contrast, lower-income households lose most heavily from the corresponding spending cut, with smaller proportionate losses for middle- and upper-middle-income households.
The result, when you look at the combined impact, is that many low-income households are ‘net losers’ from this income tax cut, when the corresponding reduction in public spending is taken into account. They are getting a big cut in public services for a pretty small gain from the tax cut.
The poorest households may well have been substantially better off if, rather than cutting income tax in the budget, George Osborne had put the money into protecting public services instead.
6) Conclusion
This report challenges the Government’s claim that the budget’s distributional impact was ‘fair’ and ‘progressive’. We show that you can only sustain that argument by ignoring the impact of the severe cuts in public spending that were also announced.
What really counts for fairness is not how families are affected by tax and benefit changes in isolation, but how they are affected by the whole package – spending cuts included.
Our analysis shows the impact of these spending cuts. The cuts will hit lower-income households much more proportionately than higher-income households. And when these cuts are taken into account, they dwarf the impact of the budget’s tax and benefit changes.
Our conclusion, then, is that the overall impact of the budget is regressive and deeply unfair.
By calculating the net impact of the budget’s income tax cut, we have also challenged the idea that cutting income tax was the optimal strategy for helping those on low incomes. Low-income households gain far less from the tax cut than they lose from the resulting fall in public spending.
Our model is based on a variety of assumptions that mean its estimates of distributional impact necessarily involve simplifications. But these are perfectly legitimate assumptions – indeed ones that the government often uses in its own fiscal analysis.
More detailed information about cuts, when it comes in the autumn, will enable us to refine this analysis. But until then, we can’t ignore the impact of tax cuts; we need some estimate of their value.
More generally, we are aiming to challenge the culture of tax and spending debates in the UK, where protagonists excitedly discuss who ‘gains’ from tax cuts, while conveniently ignoring the fact that families are hit by spending cuts.
We hope that a proper consideration of the value of public spending to households will help to rebalance these debates for the better.