A Citizen’s Income scheme’s winners and losers

In our last edition we published a research note which employed EUROMOD [note]EUROMOD was developed from the previous POLIMOD by the Institute for Social and Economic Research at the University of Essex. It employs Family Resources Survey data to calculate the actual difference in disposable income experienced by individuals when a change is made to the tax and benefits system. [/note] to calculate the gains and losses which a full range of individuals would experience if Working Tax Credits and Child Tax Credits were to be replaced with nonwithdrawable benefits. Tax Credits are only paid to people in employment, so the nonwithdrawable benefits would only have been paid to people in employment.

Here we use EUROMOD to simulate the outcomes for a genuine Citizen’s Income scheme: an unconditional and nonwithdrawable benefit for every citizen. The aim of the exercise was to test a variety of schemes for feasibility defined as follows:

The net cost of the scheme (that is, the cost of the Citizen’s Incomes, less additional Income Tax and National Insurance Contribution revenue and the money saved through the abolition of existing benefits), should be affordable. We assume that a net cost above £3bn per annum would be unaffordable in the current circumstances.

No more than 5% of individuals should suffer a loss of disposable income of more than 15%, and no more than 10% of individuals should suffer a loss of disposable income of more than 10%.

The following scheme was found to fit this definition of feasibility:

[table id=23 /]

(Because EUROMOD employs 2008 Family Resources Survey data in relation to 2009 tax and benefits regulations and levels, the state retirement age employed for this exercise is the state retirement age as it was in 2009. )

Income Tax is to be collected on all earned income above a Personal Tax Allowance of £4,000 pa as follows:

  • From £4,001 to £20,000 pa, 25%
  • From £20,001 to £40,000 pa, 35%
  • Above £40,000 pa, 45%

We have retained the Lower Earnings Limit for National Insurance Contributions, but abolished the Upper Earnings Limit. We have abolished Working Tax Credits, Child Tax Credits, Basic State Pension, contributory Jobseeker’s Allowance (but not the means-tested variety), and Child Benefit (which has been absorbed into the new Child Citizen’s Income). All other benefits and taxes have been left as they are. [note]For individuals and households claiming means-tested benefits, Citizen’s Incomes are counted as income received for the purpose of calculating the level of benefit. So, for instance, instead of Housing Benefit being withdrawn at 65% of the value of Working Tax Credits, in this scheme Housing Benefit is withdrawn at 65% of the value of Citizen’s Incomes received by the household. [/note]

The results are as follows:

[table id=26 /]


The survey covers approximately 0.1% of the total population of the UK.

Most of those individuals for whom no change occurs will be children. Their Child Benefit is ascribed to the main carer, and the Children’ Citizen’s Income would be similarly ascribed.


For the purpose of this exercise, no attempt has been made to abolish or change current means-tested benefits except for Working Tax Credits and Child Tax Credits. Because Citizen’s Incomes and Citizen’s Pensions will be in payment, an additional number of individuals should find their non-means-tested incomes to be sufficiently high to take them above the means-tested benefits levels. The number of claimants should therefore be reduced, and the amounts of means-tested benefits in payment should also be reduced.

The EUROMOD simulation, running with 2008 Family Resources Survey data and 2009 benefits regulations, shows that 11.95m adults would have been in receipt of means-tested benefits, whereas for the same data and our Citizen’s Income scheme there would only have been 7.65m in receipt. In 2009, the total means-tested benefits claimed (including Working Tax Credits, Child Tax Credits, Housing Benefit, and Council Tax Benefit) would have been £57.7bn, whereas for the Citizen’s Income scheme the total means-tested benefits claimed would have been only £26.9bn (much of which would have been Housing Benefit and Council Tax Benefit). We can therefore see that such a Citizen’s Income scheme would have reduced by more than four million the number of people suffering the disincentive effects imposed by means-tested benefits. Somewhat less calculable, but perhaps more significant, would be the number of people still in receipt of means-tested benefits who would seek additional sources of earned income in order to lift their net incomes above the level at which they would need to claim means-tested benefits. Anyone raising their earned income in the context of Working Tax Credits knows that their Working Tax Credits will be reduced and they will receive little benefit from their additional effort or additional skills. In the context of a Citizen’s Income there is no such reduction, making it rather more likely that amongst those still in receipt of means-tested benefits numerous individuals and households will do all they can to lift their incomes high enough to escape the traps that accompany means-testing.

The Gini coefficient is a measure of inequality. The Gini coefficient for disposable income in relation to the 2009 benefits regulations would have been 0.3403, whereas that for the Citizen’s Income scheme would have been 0.3304. Whilst this is not a major change, it shows that a feasible Citizen’s Income scheme might somewhat reduce inequality in disposable incomes. The considerably reduced marginal deduction rates that would be experienced automatically by four million people, and in practice by others who would choose to remove themselves from means-tested benefits, would promote increased disposable incomes in the lower earnings deciles and would thus reduce income inequality even further.

The additional cost of the Citizen’s Income scheme would have been £2.85bn, which is within the £3bn that we allowed ourselves.

This is clearly not the end of the discussion. At 4.18% the number of individuals facing a loss in disposable income of more than 15% is too high; and at 8.2% the number facing a loss of more than 10% is also too high. Some of these losses will be amongst higher earners, and some will be because a couple will now be receiving twice as much as a single person, meaning losses for single people and gains for couples.

There are three possible responses to this:

1. In the longer term, the lower marginal deduction rates that all workers will experience will mean that it will be easier than before to fill any gap in disposable income by increasing earned income. In the shorter term:

2. It should be possible to reduce these figures by adjusting Income Tax rates and the levels at which the different Citizen’s Incomes are paid. This would be at the cost of reducing the numbers of individuals seeing immediate gains in their disposable incomes. This would not be a problem. The aim of this Citizen’s Income scheme is not to increase the immediate gains experienced by a large proportion of the population. The aim is to reduce the marginal deduction rates suffered by individuals and households, and to do it in an affordable manner.

3. As Philippe Van Parijs suggests, such losses ‘do not mean that a [Citizen’s Income] is unaffordable, but that a gradual approach is required if sudden sharp falls in the disposable incomes of some households are to be avoided’. [note]Philippe Van Parijs, ‘A Basic Income for All’, pp. 3-26 in Joshua Cohen and Joel Rogers, What’s Wrong with a Free Lunch? Beacon Press, Boston, 2001, p.23[/note]

The results of this exercise suggest that by making a small number of changes to the present system it is perfectly possible to establish a genuine Citizen’s Income scheme. What is now required is a more substantial research effort to seek a Citizen’s Income scheme similar to the one outlined above but with fewer people losing more than 15% of their disposable incomes; and, just as importantly, to seek the simple transitional steps that would be required in order to get there.

Acknowledgements: We are most grateful to the Institute for Social and Economic Research at the University of Essex for the use of EUROMOD, to Professor Holly Sutherland and her colleagues for advice and tuition in the use of EUROMOD, to the European Commission for funding the development of EUROMOD, and to the UK Data Archive for the use of Family Resources Survey data