Taxing Families Fairly

August 9, 2022

Britain is an international outlier in not making proper provision in the tax system for the extra costs of child-raising carried by families. The result, as this paper shows, is that families with young children in the UK pay significantly more in income tax than comparable families in countries like Germany and France. Politicians are finally beginning to wake up to this anomaly and the issue has been raised by Liz Truss in the Conservative leadership election. But it is time that a full beam of light was shone on this bias against family formation and that is just what Philip Booth and Andrei Rogobete have done in this paper.

Related Publications


Professor Philip Booth

Professor of Finance, Public Policy and Ethics at St Mary’s University

Andrei E. Rogobete

Associate Director at The Centre for Enterprise, Markets & Ethics.

Chapter 1: Introduction

The UK’s taxation system is based on an assessment of individual income. This is not unique in the Western world, but there are many countries in which a different approach is taken. At first sight, it may seem reasonable to tax individuals on the basis of their personal income. However, to do so, when we have a system of progressive taxes, leads to highly problematic outcomes. It is also an approach that we believe is flawed philosophically. Miriam Cates MP and, more recently, both Liz Truss and Penny Mordaunt during the Conservative leadership contest, have highlighted the problems caused by our current treatment of families in the tax system and have made the case for reform.

Of course, the tax system works perfectly well for single people and our proposals in this paper would not affect them directly. However, for much of our lives, most people do not live as single persons. They live in families or households. When they do, in almost all cases, they share resources with others in the household. Indeed, this is why we measure inequality, both domestically and internationally, by looking at household incomes. It would not make sense to suggest that somebody was poor because they earned the minimum wage for ten hours a week (or earned nothing at all) whilst their spouse was paid £4m a year as a professional footballer. And yet, when we assess them for income tax, that is precisely what we assume.

It could be argued that this approach to taxation both pre-supposes and encourages a hyper-individualistic mentality. For fiscal purposes, we ignore the basic unit from which societies are built. This is a cultural and political problem which among other things leaves the UK with a bigger fertility gap (the gap between the number of children people say they want and the number they actually have) than most comparable countries.

However, there is also an economic problem. As we show below, this approach to taxation creates huge disparities between the taxation position of families that have the same income depending on whether or not their income is split evenly between members of the household. A single-earner family with a gross household income of £30,000 per annum, for example, pays far more tax than a family where two adults earn £15,000 each. This is inequitable and it creates perverse incentives for members of the household to take employment decisions based on their tax position rather than on their domestic circumstances. The system directly discriminates against households where one member takes on caring responsibilities for younger or older members of the family.

Interestingly, our welfare system tops up incomes based on an assessment of household income. The spouse of the footballer mentioned above does not receive welfare benefits simply because they have low earnings because eligibility takes into account the earnings of both spouses or both members of an unmarried couple.

When we put the tax and welfare systems together, we find that family formation is penalised – to a substantial degree. If, for example, a woman has a child and does not work whilst the father is in paid employment, if they live together or marry, her entitlement to benefits is lost whilst he continues to pay tax at the same rate as if they were single. They are, literally, better off apart. Thus, the state, through having designed an individualised tax system alongside a benefits system based on household income penalises the family as if, somehow, it is a bad thing that needs to be discouraged through taxation. Of course, all the evidence suggests the opposite. The family is important for individual wellbeing as well as having beneficial societal effects.

These questions might only be of theoretical interest if it were impossible to envisage things being done a different way. However, as we show, many countries levy taxation based on household and not individual income. Such countries include France and Germany.

In the chapters immediately following we discuss the economic and philosophical arguments for our current approach to taxation as compared with alternatives. We then follow this by modelling the cost to families with an uneven split of earnings between their main adults of the current tax system. After examining the relationship between the tax and benefits system, we examine how we can “tax families fairly” by moving to an approach along the lines of that used in France and Germany. We conclude that it is important for the UK to adopt an approach that involves taxing families on the basis of their family income and structure rather than taxation being based on the assumption that we live as individuals outside families and households.

Next chapter

Chapter 2: The unit of taxation – the family, the household or the individual?

Whether the household or the individual should be the appropriate unit for the assessment of taxation is contested in economics. We argue that the household should be the unit of taxation. This would normally be limited to those members of the household who are members of the same family. However, we will tend to use “household” and “family” interchangeably in the remainder of this paper.

The economic arguments for assessing tax on the basis of individual or family income are finely balanced when we consider the tax system alone. However, when we look at the impact of the welfare and tax systems combined, it is clear that the current tax system in the UK – which is wholly focused on taxing individual income – is flawed. The tax and welfare systems discriminate systematically against families with an unequal split of incomes and discriminate against family formation.

In the UK, individuals receive tax-free allowances and the rate at which tax is paid increases with an individual’s income. This means that a family with two adults, in which one, for example, earns £70,000 a year whilst the other takes on caring responsibilities at home pays considerably more tax than the same family which has earnings split between the two adults. In the first case, the family receives one tax-free allowance and, in the second case, two. In the first case, the single earner will pay higher rates of tax because some of their income is above the basic rate band. In the second case, because neither individual’s income is above the basic rate band, no higher rate tax will be paid. In this particular case, it should also be noted that the couple will not receive child benefit if there is a single earner with an income of £70,000 per annum whereas they would receive child benefit if their income were split between two earners. In general, in the UK tax system, two households in receipt of the same income pay very different levels of tax depending on the split of the income between the individuals. This amounts to unjust discrimination against households where incomes are unevenly split. This uneven split of incomes is most likely to occur where one of the adults undertakes caring responsibilities.

The UK system is not unique by international standards, but many countries do things differently. What are the economic arguments?

2.1 Economic considerations

A primary consideration in tax policy is the desire to achieve horizontal equity. In other words, two tax units receiving the same income should be taxed in the same way.1Musgrave, R. A. (1967). “In Defense of an Income Concept”. Harvard Law Review, 81(1), p. 44–62. If it is accepted that the unit of taxation should be the family or the household, then the current UK tax system is a long way from the principle of horizontal equity. As we shall show below, families with the same incomes, but with a different split of incomes between the adults in the households, can pay very different tax rates. But should the family be the unit of taxation? Or is the situation that pertains in the UK the most acceptable from an economic point of view.

Interestingly, when international comparisons of inequality are made, the household is the starting point for the calculations of disposable income.2Some adjustments are made for the different living costs incurred by different sized households. See: In the UK, the Office for National Statistics (ONS) evaluates inequality based on household and not individual income.3Office for National Statistics, “Household income inequality, UK: Financial year ending 2020”. Indeed, to do otherwise would be bizarre as an example easily shows. Compare two countries, A and B in households have two adults. In A, every individual earns £25,000. In B, one adult in each household earns £50,000 and the other adult, who is caring for elderly parents, earns nothing. If we measure inequality by looking at household incomes, both countries have perfect equality. If we look at individual incomes, country B would be far more unequal than country A and would have an absolute poverty rate of 50 per cent. This would not make any sense.

Though some technical adjustments are made before inequality measures are calculated, the focus of those measures is, correctly, on household income. This alone provides a strong case for taxing on the basis of family resources rather than individual resources. The latter approach presupposes that individuals live atomistically and do not share resources with others in their family.

Indeed, all other things being equal, a progressive tax system based on taxing individual income will act to increase, rather than decrease, measured household inequality according to standard measures by redistributing money from poor single-earner households to richer two-earner families.

Despite this very strong argument against an individualised tax system, there are some economic arguments that point in the other direction.

One of the most comprehensive reports on the UK tax system was the Meade Report (Meade & Institute for Fiscal Studies, 1978). Chapter 18 of this publication considered the matter of the individual and the family as a unit of taxation at some length. Overall, the report tended to favour taxation assessed at the level of the individual with some adjustments, especially in relation to investment income. But that conclusion was predicated upon the assumption that what the authors described as “home responsibility payments” would be given to non-earning spouses looking after children. Thus, an adjustment in the benefits system would be made to reflect the horizontal inequity arising within a tax system based on individual assessment. This would, of course, hugely increase welfare spending and involve taxing the resources of a family in order to then pay the same family benefits for looking after children and older people. There are many reasons why this is undesirable, but the conclusions of the Meade report cannot be used as arguments in favour of individualised taxation.

The other main comprehensive assessment of the tax system in the UK was published by the Institute for Fiscal Studies in 2011.4Mirrlees et all, (2011), “Special Issue: Special Issue on the Mirrlees Review”, The Journal of Applied Public Economics, Volume 32, Issue 3, Pg. 319-453. That report also considered the issue of the appropriate unit of taxation. It raised the question of horizontal equity and concluded that taxing households differently, depending on the split of incomes between members of the household, was not equitable.

However, Mirrlees et al made two other points that are important in any economic analysis. Firstly, they argued that using the family as a unit of taxation creates an artificial incentive to marry or cohabit (if the option of being taxed as a family unit is available to cohabitees). If two individuals have different earnings, they will be taxed less if they choose to cohabit or marry than if they remain single because they can make use of each other’s unused tax allowances. However, between median and mean earnings, this would not appear to be true in any country which uses the family as a unit of taxation if we also take into account the interaction of the tax system with the welfare system. In addition, quoting empirical work, they suggested that second earners have a higher elasticity of labour supply. This takes us into the territory of technical economics, but it means that the decisions of second earners are distorted to a greater extent by any given tax rate. According to economic theory we should therefore tax second earners less in such situations, which is exactly what tends to happen in the UK tax system. However, the extent to which second earners are taxed less is simply an accidental by-product of our individualised tax system super-imposed on the particular pattern of earnings in any given household. Our system is not somehow cleverly designed to take account of the points made by Mirrlees et al.

In addition to these arguments, it can be noted that couples with uneven incomes are more likely to be doing more “non-market”, and therefore untaxed, work.5Apps, P., Rees, R. (2018), “Optimal family taxation and income inequality”. Int Tax Public Finance, Vol.  25, p. 1093–1128. To the non-economist, it may seem strange to cite this as relevant. However, the argument can be illustrated with an example. Consider two families A and B. In family A, both partners work full-time and have similar earnings. In family B, one partner works full-time and earns considerably more than the other partner, who has a part-time job or no job at all. Combining the contributions of the two partners, family B as a whole spends less time in paid employment (including travel to work) than family A, and as a result has more time available for childcare, eldercare and domestic chores such as cooking, DIY, cleaning, and gardening. With less disposable time, family A is likely to rely more on professional carers or hire a cleaner or gardener. The taxes that are levied on those who provide these services (income tax, VAT, etc) are paid indirectly by family A, whereas the equivalent services in family B are provided by the untaxed labour of family members. An economist might say that family B has an untaxed income in kind. Individual taxation and the subsidy for professional childcare and eldercare might be seen as rough and ready ways of dealing with this anomaly.

This point is correct but, we believe, marginal.

Taking all things into consideration, Mirrlees et al. concluded as follows:

Ultimately, the choice between individual and joint taxation depends on political value judgements about how far people should be viewed as independent individuals and how far as couples. Rather than making a judgement of our own, in this book we simply take the current individual, annual assessment for taxes and joint, short-term assessment for benefits as given.6Ibid. p. 139

As is clear, we do not agree with this. However, Mirrlees et al left out another crucial argument which outweighs all others – the interaction between the tax and benefits system.

In the United Kingdom, welfare payments are based on family income. This means that, as family income rises, benefit payments are reduced. This is reasonable. Welfare systems are designed to redistribute money from rich families to poor families. However, if an individual who is earning money forms a household with somebody who is not, the couple lose significant amounts in welfare benefits. In our tax system, which uses the individual as a unit of taxation, the couple do not gain through lower taxation when forming a household. This is quite unlike in many other countries where, if a non-earner and an earner form a household, the earner would receive two tax-free allowances. The reduction in the couple’s tax bill then helps to compensate for the loss of welfare benefits. There is symmetry between the welfare and tax systems in such cases.

The International Federation for Family Development (IFFD) found overwhelming evidence for the positive outcomes that come from family stability. These include: higher life expectancy, lower risks of mental illness and drug addiction, higher academic achievement, increased discipline in fulfilling legal duties and social contributions to society.7IFFD, (2017), “The Crucial Role of Families”, It can genuinely be said that, in the UK, we subsidise people to live apart – almost as if the creation of families brings what economists call “negative externalities”. It seems as if politicians believe that family formation is damaging and needs to be financially penalised in the same way as the consumption of cigarettes or sugar.

Contrary to the suggestion of Mirrlees et al, a tax system based on the individual as a unit of assessment discriminates against, rather than artificially encourages, marriage and cohabitation.

The last major work on couple penalties in the tax and benefit system was by Adam and Brewer (2010).8Stuart Adam, Mike Brewer, (2010), “Couple penalties and premiums in the UK tax and benefit system”, Institute for Fiscal Studies. They concluded: “68% of existing couples in the UK face a penalty in the 2010–11 tax and benefit system, 27% face neutrality and 4% face a premium…The sum of all couple penalties amounts to £34.7 billion a year…”.9Ibid. pg. 4. The couple penalties are concentrated in families with children: “By family type, almost all (95% of) couples with children have a couple penalty, as do 81% of pensioner couples and 41% of working-age couples without children.” Their conclusion is similar to ours that the penalty on family formation arises from: “…paying out benefits and tax credits according to family circumstances but levying income tax according to an individual’s income”.10Ibid. pg. 5, 75.

The reversing of Mirrlees’ argument in this way, combined with the horizontal equity argument, points strongly in the direction of the use of the family as the unit of taxation. Of course, the other economic arguments in favour of individual taxation still stand, but they seem to be of marginal importance.

Given this, it is extremely difficult to see how the views of the former Deputy Prime Minister, Nick Clegg, on the small transferable tax allowance that exists in the UK can be justified. He argued against it saying that there was a limit on what the state “should seek to do in organising people’s private relationships” and that “We can all agree that strong relationships between parents are important, but not agree that the state should use the tax system to encourage a particular family form”.11Simon Alford, (Dec. 18th 2011), “Clegg attacks marriage tax break plans”, The Times. This view is widely held in Westminster, but it suggests a stunning lack of familiarity with the UK tax and benefit system for which he was responsible.

2.2 The family as the basic unit of social organisation

There are, of course, arguments that transcend or complement those arising from economic considerations alone.

In most religious traditions, the family is the basic unit from which society is built. For example, in his 1963 encyclical entitled Pacem in terris, (“Peace on Earth”), Pope St. John XXIII wrote extensively on the role of the state, the rights and responsibilities of individuals and the importance of the family as a unit of social organisation. Here he argued:

The family, founded upon marriage freely contracted, one and indissoluble, must be regarded as the natural, primary cell of human society. The interests of the family, therefore, must be taken very specially into consideration in social and economic affairs, as well as in the spheres of faith and morals. For all of these have to do with strengthening the family and assisting it in the fulfilment of its mission.12“Pacem In Terris” – Encyclical Of Pope John XXIII on Establishing Universal Peace In Truth, Justice, Charity, And Liberty, (April 11, 1963),

The family is therefore the “primary cell” that forms and shapes society and has always been treated as such in the Christian tradition. This has been reflected in government policy down the ages. However, this is also a position that is widely accepted by people regardless of religious belief. Taking a more evolutionary perspective which does not rely on a Christian anthropology, F. A. Hayek has written about the central place of the family as the integral building block of society. For example, in The Constitution of Liberty he states: “[S]ociety is made up as much of families as of individuals and that the transmission of the heritage of civilisation within the family is as important a tool in man’s striving towards better things as is the heredity of beneficial physical attributes.”13Hayek (1960), p. 90.

Hayek famously wrote about social justice (by which he meant distributive justice) suggesting that there could be no general agreement upon a set of rules which could be applied in order to ensure a just distribution of incomes within society without undermining the rules of just conduct on which a free society is built. At the same time, Hayek wrote eloquently in his volume, The Mirage of Social Justice14Hayek (1976)., about how principles of distributive justice could be developed and applied at the level of the small group. We would argue that the first, and perhaps most important, level at which distributive justice is carried out is within the household. Those earning money within a household share their resources with others in a household, including non-earning spouses, children and, sometimes, elderly, relatives. Mueller (2014) argues that the idea of “distribution” has been neglected in economics. This is probably true when it comes to the distribution of resources that takes place below the level of the state. The family is the most fundamental level at which economic resources are distributed.

Progressive taxation works under the assumption that those with greater resources should contribute more. To base the resources available for taxing on an individual assessment rather than an assessment at the level of the family simply ignores reality. Of course, many households have only one member and so this debate is irrelevant for them. But, when adults live together (whether married or not), they share resources. Whether somebody is rich or poor depends not only on their individual earnings but on the earnings of others in their household. Philosophically, it makes sense to levy taxes based on the capacity of the family to pay.

To not base tax assessments on family income also distorts decision-making. Given the very strong bias against single-earner families in the tax system that is discussed below, we are artificially encouraging families to become two-earner families and/or to organise their working patterns so that they have a roughly even split of incomes between the family members. When we also consider the discrimination against family formation arising from the interaction of the tax and welfare systems, our whole fiscal approach can be described as being explicitly anti-family. Of course, other factors may be to blame too, but it is worth noting that the UK has one of the highest rates of prevalence of single-parent families in the developed world.

Next chapter

Chapter 3: From families to individuals as the unit of taxation – a brief history

It might be thought that the conclusion of the previous section is “quaint” or “old-fashioned”. The UK’s individualised tax system seems deeply embedded. There has been some discussion of alternatives to our current tax system by organisations such as Care15Leonard Beighton, Don Draper and Alistair Pearson, (2018), “The Taxation of Families”, Care – Tax & the Family,   and there is a great deal of research produced by Tax and the Family16Tax & the Family Reports, As noted above, the Institute for Fiscal Studies has raised the question of couple penalties. Patricia Morgan has also undertaken research on this issue, concluding in Morgan (2007) that: “families should be allowed to retain resources on a par with those available to individuals without dependants, probably through transferring tax-free entitlements”.17Patricia Morgan, (2017), “The War between the State and the Family: How Government Divides and Impoverishes”, Abingdon: Taylor & Francis, p. 148-149. Williams (2019) goes even further in arguing that arguing:

By forcing a married couple to be treated as two individuals, the UK tax system fails as good public policy. […] our tax system is philosophically incoherent. Stable families and stable marriages are the bedrock of a strong and flourishing society.18Jonathan Williams, (June 18th, 2019), “Taxing Families in the UK”, Institute for Fiscal Studies,

As noted above, there was opposition to the introduction of the marriage allowance, even within the coalition government of 2010-2015. But, taken as a whole, even that government’s policy was incoherent. The same government that brought in the small marriage allowance also brought in the withdrawal of child benefit in such a way that it discriminated strongly against single-earner families (see below).

Many other developed countries levy tax on the basis of family income. For example, in France and Germany, as will be described below, household incomes are, in effect, pooled and then divided between household members before calculating tax due.

Indeed, the UK, in some senses, arrived at our current place by accident rather than by design The treatment of the family in our current tax system traces its roots to the early 1980s when then Chancellor of the Exchequer Nigel Lawson sought to implement a tax system that would be “neutral and fair” to the individual.19Tax & the Family, (2018), “An Interview with Lord Lawson about Independent Taxation”, YouTube, But there was a caveat. In his memoirs Lawson argued that, “…after careful study, I came to the firm conclusion that this should be a system of independent taxation, with the allowances freely transferable between husband and wife.”20Nigel Lawson, (1992), “The View from No.11: Memoirs of a Tory Radical”, London: Bantam Press, p. 882. The premise was rather straightforward:

“Everyone, man or woman, married or single, would have the same standard allowance. But if either a wife or a husband were not able to make full use of their allowance, the unused portion could be transferred, if they so wished, to their partner. […] It would end the present discrimination against the family where the wife feels it right to stay at home, which increasingly nowadays means discrimination against the family with young children.”21Ibid. p 883

This was never achieved. Lawson describes how Prime Minister Thatcher, “…did not like the idea at all”. Her reaction came as a surprise since she undoubtedly valued the family. In practice however, Lawson points out that “…she strongly identified with two-earner couples”.22Ibid. p 882

Re-confirmed as Chancellor following the 1987 general election, Lawson ultimately settled for what he describes as a “halfway house” with independent taxation but without fully transferable allowances. Families would however benefit from a ‘Married Couple’s Allowance’ which originally represented the difference between the single person’s allowance and the married man’s allowance in the old system.23Ibid. p 885 Even at the beginning, this was a poor substitute for transferable allowances, but it was, in any case, gradually withdrawn. Bizarrely, it was kept for pensioners.

A very limited marriage allowance was brought back by the coalition government, as noted above. For the 2019-20 tax year, the Marriage Allowance allows single earner couples or couples where one person earns less than £12,500 per year to transfer a £1,250 additional tax allowance to their partner. This would result in a maximum annual tax saving of up to £250.24GOV.UK, (2021), “Marriage Allowance”, The mechanism is complex and the total maximum benefit is small. It is also withdrawn as incomes rise. This was accepted by David Cameron who said: “This policy isn’t about the money but about the message that people who make a lasting commitment should be recognised in some way.”25David Cameron, (14 January 2019), “House of Commons Briefing Paper Number 4392: Tax, marriage & transferable allowances”, p. 32.

This seems to be a rather weak statement: surely, the strength of the “message” depends on the extent of the financial benefit. But the idea has been packaged in the wrong way from the beginning. The marriage allowance has been promoted as a way of rewarding marriage rather than a way of gradually ending discrimination against households where the split of earnings between individuals within households is uneven. Also, the marriage allowance cannot be said to have been successful. The take-up has been low: in 2019, fewer than half of eligible couples made use of the allowance.26Sam Bromley, (26 June 2019), “Revised HMRC figures reveal more than 2 million people are missing out on this tax break”, Simply Business, Presumably, this is due to its complexity and relatively low value.

So, what are the financial impacts of our tax system on families with different earnings profiles?

Next chapter

Chapter 4: Taxing families unfairly – the unequal treatment of households in the UK tax system

Here we quantify the extent of the variation of the tax burden between families with two earners and those with one primary earner. We look at the total tax paid in different situations and the average tax rates (ATR) of households where earnings are split in different ways between two main adults. We consider both income tax and national insurance and also examine the effect of child benefit withdrawal.  In this section, and in the following sections, we have used tax and benefit rates and bands that were correct in April 2020. We considered using more up-to-date figures but in the intervening two years, the rates and bands have continually changed, even within fiscal years, which would make some calculations impossible. Almost every change, including the forthcoming increase in the level of earning before National Insurance Contributions become due, have or will worsen the problems we discuss. The figures below illustrate the problem with an acceptable degree of accuracy.

4.1 Income Tax

Figure 1 shows how the average tax rate changes for families with different levels of household incomes (shown by different coloured lines) as the split of income becomes more uneven. At the left-hand side of the figure, the average tax rate is that which prevails where there is a 50:50 split of income. The right-hand side represents a single-earner family. There is then every possible split in between those two extremes.

Figure 1: Average Tax Rate (ATR) at various income levels and earning compositions

The following specimen examples are worth noting. Median household earnings are about £30,000 per annum and so examples have been chosen at representative points below and above median earnings.

  • The amount of income tax paid by a household earning £20,000 annually varies from £0 to £1,500 depending on the composition of household earnings, with an effective average income tax rate varying from 0–7.5%.
  • The amount of Income Tax paid by a household earning £30,000 annually varies from £1,000 to £3,500 depending on the composition of household earnings, with an effective average income tax rate varying from of 3.3–11.7%.
  • The amount of Income Tax paid by a household earning £70,000 annually varies from £9,000 to £15,000 depending on the composition of household earnings, with an effective average income tax rate of 12.9-22.1%.
  • The amount of Income Tax paid by a household earning £140,000 annually varies from £31,000 to £48,500 depending on the composition of household earnings, with an effective average income tax rate of 22.1-34.6%.

In other words, the income tax system places single-earner families in an unfavourable tax position regardless of income level: they are consistently paying the highest ATRs, often higher by a significant margin. If we take the approximate median household annual income of £30,000, we can see (as shown in the examples), that a single-earner family will have a disposable income of £2,500 less as a result of their higher tax bill compared with a household where both spouses are earners. This would require the single-earner household to earn an additional £3,125 in order to have the same disposable income as the dual-earner household.

The situation becomes more dramatic as we move up the earnings scale. We should remember that couples higher up the income scale relative to median earnings should not necessarily be regarded as well off, especially if they live in areas with high housing costs. The median household income is much higher for couples who are of the age where most people have children, when decisions about which parents go out to work have to be taken, but resources are also often tighter.27Median weekly pay, for example, is 27 per cent higher for people in their 40s as compared with people in their 20s. See: “Average earnings by age and region”, House of Commons, The aggregate median income is based on the distribution of incomes across society as a whole, including those of younger and older ages who would be expected to have less income, on average, than families with earners in their late 30 and 40s.

At a family income level of £70,000, a single-earner family will pay £6,000 more in tax than a dual-earner family with an equal split of earnings (i.e., £35,000 per annum each). The single-earner family will have to earn an additional £10,000 to have the same disposable income as the dual-earner family – that is, the single-earner family requires earnings that are 14% higher.

At a family income level of £140,000, a single-earner family will pay £17,500 more in tax than the dual-earner family who earn £70,000 each. The single-earner family needs to earn an additional £29,000 to have the same disposable income as the dual-earner family: in other words, 20% more.

It should be noted that these are conservative estimates as they ignore national insurance contributions and the withdrawal of child benefit the impact of which are examined in later sections.

4.2 Income Tax and Marriage Allowance

The impact of the Marriage Allowance (MA) on household finances is largely insignificant. Regardless of the level of total income, the maximum tax saving a family can claim using the MA is £250 per annum. It is no surprise, therefore, that the resulting reduction in household ATR is also negligible.

The following examples illustrate the point:

  • The MA reduces the effective average income tax rate from 7.5% to 6.3% for a single-earner couple on £20,000 per annum. The dual-earner couple with a 50-50 income split will stay pay no income tax at all and so has an ATR of zero.
  • For a single-earner couple earning £30,000, the MA reduces the effective average income tax rate by just 0.9% from 11.7% to 10.8%. The effective ATR for the dual-earner earner with a 50-50 split remains at just 3.3%.

Figure 2 illustrates the ATR of households when the MA is taken into consideration. As noted above, the reduction in ATR resulting from the marriage allowance is tiny, even when its effect is at its greatest. The reduction in ATR from the MA continues to decrease in percentage terms as the household income increases. It is notable from this figure that, even after allowing for the MA, a single-earner couple earning £50,000 a year faces approximately the same average tax rate as a dual-earner couple earning £90,000.

Figure 2: Average Tax Rate (ATR) with Marriage Allowance (MA) at various income levels and earning compositions

4.3 Income Tax, Marriage Allowance, Child Benefit Tax Charge and National Insurance Contributions

So far, we have focused on the income tax system and ignored National Insurance Contributions (NICs). The situation with regard to national insurance is more complex conceptually because a contribution record provides an entitlement to benefits – most importantly a state pension in retirement.

If the benefit received from the payment of NICs were proportional to the amount of NICs paid, we might not be worried about the differences in the incidence of the national insurance system on different households: if more contributions were paid, the benefit received would also be greater. However, in the UK system, national insurance is, from the financial perspective, more like a standard tax. Individuals become eligible for accrual of the full state pension as soon as they enter the national insurance system28Like many aspects of the UK tax system, there are unnecessary complexities. There is a small income band in which workers earn benefits without paying contributions.. NICs are then paid, with no entitlement to further benefits, until earnings reach approximately £50,000. They then reduce dramatically for the employee. This leads to a particular additional problem of discrimination against single-earner households. A single-earner household with earnings of £50,000 per annum, for example, will pay more national insurance than a dual-earner household where two adults each earn £25,000. In total, they will pay about £800 more. However, whereas the dual-earner household will receive entitlement to two state pensions, the single-earner household will be entitled to only one. So, the dual-earner family pays less in NICs and receives twice as much in terms of state pension entitlement. This is mitigated somewhat by the fact that, in some circumstances, they can receive state pension entitlement when they have caring responsibilities.

In the calculations below, we have only included employee national insurance contributions. We have ignored employer NICs, the whole or part of the burden which is likely to be borne by employees. The figures presented therefore understate the discrimination against single-earner families.

An additional consideration arises from the Child Benefit Tax Charge (CBTC), also known as the ‘High Income Child Benefit Tax Charge’. This is a tax charge placed on incomes above £50,000 per annum for households in particular circumstances. The tax charge arises from the desire by the government to withdraw child benefit from those on higher incomes: though, as we have noted, a single-earner family earning £50,000 a year in an area of high housing costs should not necessarily be regarded as “high income”.

Its operation seems to have been designed to impose a discriminatory burden on single-earner families. If any earner within a family earns more than £50,000 per annum, the child benefit is withdrawn, even if the child benefit was paid to a non-earner (for example, the mother). The withdrawal takes place over the following £10,000 of income until all the child benefit has been withdrawn once the individual is earning £60,000. The cost of this depends on the number of children, but it can be enormous and can add substantially to both the marginal and average tax rate of a household that has one or more individuals earning over £50,000 per annum. For example, a family with three children with one member of the household earning £60,000 will lose their £2,556 child benefit as their earnings progress from £50,000 to £60,000.

There is no principle of taxation that can justify this policy. A dual-income household with an income of £100,000 evenly split between two earners will not be affected at all by the charge, whereas a single-earner family with an income of £60,000 and three children would lose a total of £2,566. This would amount to 4.3% of gross income and 5.9% of net income.

It is worth noting that, even for people on average earnings, the operation of our tax and benefits system is such that the disposable income of families with children is hardly bigger at all than those without. The UK neither reduces the tax burden nor increases welfare benefits as much as in most other countries for families with children. The operation of the CBTC makes things even worse higher up the income scale.

Figure 3 shows the ATR taking into account income tax and national insurance as well as the Marriage Allowance and Child Benefit Tax Charge for households with different earnings’ structures and at different levels of total earnings.

  • The amount of the Income Tax and National Insurance Contributions paid by a household earning £30,000 varies from £1,502 to £4,894 depending on the household earnings’ composition, with an effective average tax rate varying between 5% and 16%. A single-earner household pays £3,392 more in tax than a dual-earner household on the same level of income. The single-earner family would have to earn an additional £4,988 per annum to have the same disposable income as the dual-earner family with a 50-50 split of incomes.
  • The amount of the Income Tax and National Insurance Contributions paid by a household earning £70,000 varies from £14,302 to £22,237 depending on the household earnings’ composition, with an effective average tax rate varying between 20%–32%. This would result in the single-earner family paying £7,935 more in tax compared with the dual-earner household on the same level of income. The single-earner family would have to earn an additional £13,681 to have the same disposable income as the dual-earner family with a 50-50 split of incomes. In other words, the single-earner family would have to earn 20% more to have the same disposable income as the family with an equal split of earnings.
  • The amount of the Income Tax and National Insurance Contributions paid by a household earning £140,000 varies from £42,997 to £56,637 depending on the household earnings’ composition, with an effective average tax rate varying between 31% and 40%. This would result in the single-earner family paying £13,640 more in tax compared to dual-earner household on the same level of income. The single-earner family would have to earn about £23,500 more to have the same disposable income as the dual-earner family with a 50-50 income split.

It should be remembered that, in all these cases, the dual-earner family is more likely to accrue two state pensions than the single-earner family. However, it should also be noted that, at very high levels of earnings, the dual-earner family will pay more in National Insurance Contributions in total than the single-earner family.

4.4 Conclusion

The analysis so far has shown how the tax system in the UK systematically discriminates against families where there is an uneven split of earnings between the two members of a couple. This discrimination is difficult to justify and is a consequence of a tax system that focuses obsessively on the individual rather than the household. This is despite the fact that government inequality measures and policies are focused at the household level.

It should be noted that an unequal split of earnings between the members of a couple can arise for a variety of reasons. It could arise as a result of unimpeded choice whereby the couple decides that one of its members will not work or only work part-time in order to take on caring responsibilities. Such choices can, of course, be strongly influenced by family circumstances in which one member of the couple may have little choice but to take on caring responsibilities. Differences between earnings within a couple could also arise because of accident or illness. Or they could arise because of differences in skills levels, training or experience between the two members of a couple. Our contention is that, whatever the cause, two households with the same income should face approximately the same tax bill. We emphasise “approximately” because there might be a desire to make some allowance for the difference in costs per person of running a multi-member household. The starting position, however, should be one of treating households with the same income equally as happens in countries such as Germany and France.

This conclusion is reinforced when we consider how the tax system interacts with the benefit system. We then see that our individualised tax system, in combination with our benefits system, discriminates not just against single-earner families but against family formation itself.

Next chapter

Chapter 5: The Interrelationship between Household Composition and Universal Credit

Chapter two briefly discussed the incoherence of assessing income tax on an individual basis whilst measuring inequality and assessing benefits on a household basis. The interaction of the tax and welfare systems in the UK creates a range of problems. Because Universal Credit (the basic welfare benefit in the UK) is based on household and not individual income, single people can be financially better off by remaining single. If the wage-earner forms a household with a non-wage-earner (especially with children), then the non-wage-earner is likely to lose their welfare benefits whilst the wage earner would pay exactly the same amount of tax (with a small saving if they are eligible for the marriage allowance). This leads to a benefits system that disincentivises couples living together or marrying – or, at least declaring that they are living together. This is especially so if they have children. Even if it is thought that financial incentives do not affect behaviour, this situation raises questions about fairness and the impact of our fiscal system on culture.

The range of potential scenarios for benefit entitlements and tax obligations is enormous given that the systems interact with each other. Furthermore, welfare benefits are dependent upon expenses faced by households. We have therefore chosen a few examples to illustrate the problems. They are unlikely to illustrate the most egregious situations.

Table 1 shows how much a couple stand to lose financially under Universal Credit (UC) if they decide to live together rather than separately. It should be noted that the complexities go beyond the variation of benefits with household situation. Each household is assessed on a case-by-case basis with the amount of UC that each claimant will get being based on an ‘assessment period’. This represents a rolling one-month period that starts on the day that the first claim for UC is made.29,21%20November%20to%2020%20December. Therefore, there will be practical variations in relation to the timing of receipts when a family’s situation changes. The calculations are thus representative as an illustrative average.

Table 1 assumes an hourly minimum wage rate of £8.91 and a full-time and part-time working week of 37h and 18.5h respectively. Our household assumes two parents and two children under the age of 5. The person receiving UC and holding childcare responsibilities in each case is “Partner A”. The financial data was generated using a UK government recommended benefits calculator.30Turn2us – Figures representative as of 15th Oct. 2021.

Tax and national insurance was applied to earnings and couples were assumed to benefit from the Married Couples Allowance where this was available to them.

Table 1: Single vs Couple Total Income (after Tax/ incl. UC)

Partner A income Partner B income Monthly Income when living: Loss in Total Income (%) if living together, rather than separately Loss in Total Universal Credit (%) if living together, rather than separately
Together Separately
£0 £330/week (full time work at the minimum wage, £8.91*37h) £1,707.53 £2,000.98 -14.66% -37.0%
£0 £0 (unemployed and looking for work) £984.06 £1,123.81


-12.4% -12.4%
£165/week (part-time work at the minimum wage, £8.91*18.5h) £330/week (full time work at the minimum wage, £8.91*37h) £1,915.86 £2,535.79 -24.44% -92.1%


£165/week (part-time work at the minimum wage, £8.91*18.5h) £0 £1518.09 £1657.84


-8.4% -14.0%

Table 1 presents four scenarios:

  1. Partner A without income, living together with and living separately from partner B who works full time earning a minimum wage.
  2. Partner A without income, living together with and living separately from partner B who is unemployed.
  3. Partner A working part time, living together with and living separately from partner B who works full time earning a minimum wage.
  4. Partner A working part time, living together with and living separately from partner B who is unemployed.

If an individual who has no income choses to form a household with a partner who works on the minimum wage, between them they will both see a reduction in their total income of nearly 15%. The worst-case scenario in this illustration arises where one partner works part time and the other full time. This arises because they lose a significant amount of universal credit whilst paying a significant amount of tax as a result of the uneven split of earnings between the partners.

In one sense, the basic problem here is simple. However, as the discussion of Nick Clegg’s position highlights, the explanation is subtle and seems, therefore, to be misunderstood even by Deputy Prime Ministers. The welfare system works on the basis that families exist to share resources. It therefore provides resources to families that have insufficient resources. The welfare system does not provide resources to individuals who have no income within well-off families. It therefore takes away resources when somebody with no income forms a household with a partner who has an income. These kinds of incentives are natural in any welfare system. However, in our individualised tax system, tax is not charged according to family resources but according to the incomes of individuals. This means that if you put an earner into a household with a non-earner, the earner is taxed as if he continues not to share his income with the non-earner and any children – but they do not receive any welfare benefits.

A report published by The Health Foundation found that in July-August 2020 over 200,000 people were ineligible for UC solely due to their partner’s earnings and a further 200,000 were ineligible due to their own savings.31Geiger, BB; Scullion, L; Summers, K; Martin, P; Lawler, C; Edmiston, D; Gibbons, A; Ingold, I; Karagiannaki, E; Robertshaw, R and de Vries, R (2020), “Should social security reach further? A study of those not claiming benefits at the start of the COVID-19 pandemic”, Welfare at a (Social) Distance Project Report. The report concluded that this placed additional financial pressure and difficulty on those choosing to live together:

“…there is significant financial strain among those who were ineligible for benefits (UC and new style Job Seekers Allowance): much more than among the general public. […] This financial strain was particularly acute amongst those ineligible for UC due to partner earnings, and amongst those experiencing income shocks who were ineligible for UC and contributory benefits. Many (though not all) felt that it was unfair that they were not eligible.”32Ibid. p. 28

It is reasonable, of course, for the state to only assist households who have no income. The welfare system should not be designed to provide support for individuals who have not income who live in households who are earning above subsistence levels. In this sense, a welfare system is always likely to penalise household formation at least to some extent. However, if the necessity for welfare payments is determined by household income, how can it be argued that the ability to pay taxes is not also determined by household income?

One solution to achieving some sense of financial parity between couples and individuals requires structural change in the criteria under which UC is paid out to claimants. The Institute for Fiscal Studies (IFS) suggested 2010 that “Existing couple penalties could be reduced, and some couple premiums created, by increasing benefits and tax credits for couples or cutting them for single adults.”33Institute for Fiscal Studies, (2010), “Press Release: Couple ‘penalties’ in tax and benefit system are widespread, but almost impossible to eliminate”, Regardless of whether couple benefits are increased or single benefits are decreased, the current discrepancy between couples contributes to the bias against family formation.

Next chapter

Chapter 6: Taxing families fairly, the example of Germany and France

As we have discussed earlier, our analysis does not imply trying to design a tax system that is based on some 1950s vision of the typical family. The modern reality is that people live form households and different adults in the household contribute differently in terms of paid and unpaid work. Our tax system recognised this reality until recently and the tax system of many European countries still does today.

The table below showcases a few examples of countries where joint taxation is possible and family composition is taken into consideration when determining tax liabilities.

Country Tax System in relation to Family Composition
France Under the principle of “Quotient Familial” – Family taxation is based on a “quota”, or number of adults and children within the household.
Germany Based on the principle of ‘income splitting’: the tax of a married couple is determined by taxing half of their combined incomes and then doubling the amount to result in the total tax payable.
US Two spouses (as a married couple) can file jointly on their combined household income. Tax relief is then calculated based on income and the number of dependants.
Poland Joint Tax settlement between spouses is permitted. Tax deductions are calculated based on family composition (as of 2021 the tax relief is approx. £200 for the first and second child, £350 for the third child, and £470 for fourth and each subsequent child in a family).
Ireland All legally married couples are eligible for Joint Assessment. Under joint assessment couples can reduce their tax liabilities to up to €5,050 per annum. It allows couples to split their tax credits and rate band between both parties. If only one partner earns a taxable income, Revenue assigns all tax credits and the standard rate cut-off points to that person. In Ireland it is almost always more advantageous to submit a joint assessment.
Spain Two spouses can file jointly and total tax liabilities are calculated on income level and family composition, with tax breaks and/or allowances for each dependant under the age of 25.
Canada A Family Tax Cut calculation (FTC) if offered to spouses with minors. This is based on the principle of income splitting and is offered in the form of a tax credit on total household income.
Switzerland The tax rate applicable to a married couple in a Swiss registered partnership is the rate applicable to 50% of their combined income (so-called ‘splitting’). Single-earner households can benefit from reduced tax rates.
Belgium Spouses may file tax returns jointly. Tax breaks are dependant on income distribution. Single-earner households benefit from what is called the “marital deduction”, which is up to €11,170 (in 2021).

Before discussing those alternatives to our current tax system, we will look briefly at a proposal made by a former Treasury minister and Cabinet minister in the Labour government led by Tony Blair

6.1 Taxing families fairly by reforming welfare

One way of addressing the problems discussed in Chapter 4 was proposed by Rt Hon Ruth Kelly, in a pamphlet for the Fawcett Society in 2000, published not long before she became Economic Secretary to the Treasury in the Blair Government34See:

Recognising the financial penalties from marriage and household formation, Kelly suggested that, when an individual in receipt of welfare benefits forms a household with an earner, the partners incomes are added together and divided by two. They are then treated as individuals as far as the calculation of welfare benefits are concerned. To illustrate the points at the extremes, suppose an individual with earnings of £100,000 married an individual with no earnings. The welfare benefits of both parties would be based on an income of £50,000. Therefore, neither would receive any benefits. On the other hand, if an individual earning the minimum wage (£17,160 in our example above) married an individual with no income, they would both apply for welfare benefits based on the assumption that they were both single-person households earning £8,580).

This proposal would have the merit of removing some of the disincentives towards family formation that exist in the UK fiscal system. It would also work within the existing framework and perhaps be more palatable to politicians. However, it has a serious drawback. It would increase the total amount of money spent on welfare and increase welfare churn by which families simultaneously pay taxes and receive welfare payments. Our main proposal therefore involves a radical change to the tax system to leave resources in the hands of families.

6.2 Taxing families fairly – the examples of Germany and France

The German tax system makes provisions for families by a principle of ‘income splitting’. Under this principle the tax of a married couple is determined by taxing half of their combined incomes and then doubling the amount to result in the total tax payable. Given the German progressive system of taxation, this method would result in a lower amount of total tax for the couple than if they were taxed individually.35Viktor Steiner, Katharina Wrohlich, (2004), “Household Taxation, Income Splitting and Labour Supply Incentives – A Microsimulation Study for Germany”, German Institute of Economic Research, DIW Discussion Papers No. 421. This system means that households with the same income pay approximately the same amount of tax regardless of the split of income between individuals within a household.

In addition, there are also special provisions for children such as child tax free allowances. Though it is beyond the subject of this paper, this would be a useful evolution of the system we propose here. If cash benefits for children were replaced with additional tax allowances, it would reduce “churning” by which families both pay tax and receive welfare benefits for their children36See Niemietz (2010) for an explanation of how such a system could work. If a household’s income was less than the total of the tax allowances, the household could then receive a benefit payment in much the same way as universal credit works. However, the approach would take people out of the universal credit system overall.. The German system has rather extensive measures in place to financially help and promote family formation.

Table 1 shows how much income tax households would pay if the German principle of income splitting were applied in the UK. A single-earner household on a yearly income of £30,000 would pay £1,000 in annual tax given these assumptions. Under the current UK system, that same family pays £3,250 in income tax.

At an annual income of £70,000, a single-earner household would pay £15,500 in tax under the current UK system compared with £9,000 if the principles underlying the German system were applied. This means that the family pays 172% more tax than it would under the German system. The ATR consequently jumps from 13% under the German system to 22% under the current UK system.

Table 2 shows households with a 75/25% income composition. A household with an annual income of £30,000 currently pays £1,750 in tax with an ATR of 6%. Under the German principle of income splitting, that same household would pay £1,000 in tax with an ATR of 3%.

At £90,000 per annum, a household with a 75/25% income composition currently pays £16,500 in income tax with an ATR of 18%. Under the German system the total tax owed would be £13,000 with an ATR of 14%.

The French system is slightly more complex than the German approach. In France family taxation is based on a “quota”, or number of adults and children within the household. The term in French is “quotient familial”. It considers the entire composition of the household to determine the number of dependants, and thus the total amount of tax, (2021), “Guide to French Income Tax”

In a nutshell, each adult is the equivalent of one unit or ‘share’. The first two child dependants are counted as half a unit each, whilst the third (or subsequent) children are counted as 1 unit each. The total household income is then divided by the household ‘quota’ number which, for instance, in a four-person family with two adults and two children would be three. The amount of tax is then calculated based on the tax bracket that the divided sum falls into, and total is multiplied by the household quota (which in this case would be three). The result is a lower overall tax bill for the household that is primarily driven and determined by the number of dependants – not just the number of earners and split of earnings between them.

Once again, we would favour an approach recognising the cost of raising children that, where possible, cut tax bills rather than involved government spending. However, that is a separate issue not considered further here. As far as the adults are concerned, the basic principle is clear: the total amount of tax paid is not affected by the split of earnings between the adults in the family.

Table 3 and Table 4 illustrate how much tax would be paid by UK households if the French principle of quotient familial were applied (but not using French tax rates or bands). Once again, we do not take into account national insurance contributions or the child benefit tax charge. We do, however, assume that the family has two children.

Table 3 illustrates the situation of single earner households with a quota number of three (i.e., four individuals: two parents, two children), at various levels of income. Here we can see how, at an annual income of £30,000, a UK household currently pays £3,250 in tax with an ATR of 11%. Under the French principle of quotient familial, that same household would pay no tax (an ATR of 0%).

At £40,000, a UK household currently pays £5,250 in annual tax with an ATR of 13%. Under the French system, that household would pay £500 in income tax and have an ATR of 1% – resulting in a significant tax reduction of £4,750.

A UK household earning £70,000 currently pays £15,500 in annual tax with an ATR of 22%. Under the French system, that household would pay £6,500 in income tax and have an ATR of 9% – resulting in a tax saving of £9,000.

A household earnings composition of 75/25% also reveals significant tax savings under the French system compared with the current UK system for a family of four. In Table 4 we see how a family earning £30,000 per annum would pay no income tax under the quotient familial principle compared with £1,750 under the current UK system.

At an income of £50,000 per annum the tax bill would be halved from £5,000 to £2,500 by adopting the French system. Even at comparatively higher levels of income the tax saving for the household remains notable. At £90,000 the tax bill would be reduced from £16,500 to £10,500 – saving the family a total of £6,000 per annum.

It is important to note that these calculations are illustrative. If the tax system were changed in the way we are proposing, it is likely that revenue would be reduced so that, for a given level of government spending and if other taxes remained the same, income tax rates or thresholds would have to be adjusted in a way which would increase the burden on some families and single people and offset some of the savings we have indicated families may benefit from. However, the following should also be noted:

  • Tax rates would not need to increase and thresholds reduced to German and French levels as these are also higher because the government spends more in those countries.
  • There would almost certainly be behavioral changes arising from reduced fiscal costs of family formation which are likely to reduce welfare costs and increase tax revenues. Modelling this is beyond the scope of this paper and, indeed, likely to be beyond the capability of Treasury models.
  • In both the French and German systems (though it is more explicit in the French system), the tax burden is reduced for families with children. If this were to be adopted in the UK (which is not our main proposal), there would be a number of benefits. The tax burden on families with children would be reduced and there could therefore be a corresponding a reduction in welfare payments to families with children who are also paying tax. Indeed, as proposed by Niemietz (2010) the tax and welfare systems could be integrated for such families so that, if the income of a family was below the level of tax-free income after taking into account the allowance in respect of both adults and children, a cash benefit would be paid which would be withdrawn as income rose. This would, in effect, involve some integration of the tax and universal credit system.
Next chapter

Chapter 7: “Costs” of Reforming the Taxation of Families

Whenever proposals such as this are put forward, analysts, and especially the Treasury, talk about their “costs”. We should state at the outset that there are no costs to the Treasury of this proposal. The language of costs pre-supposes one of two things. The implicit assumption is being made either that all money belongs to the Treasury so that, if it takes less in taxes, it is a “cost” to the government and/or that our proposed tax system for families is conceptually wrong. For reasons we have explained, taxing families on the basis of aggregated income with the tax allowances also being aggregated is the conceptually and practically correct way to levy taxes. A move towards our proposed system would reduce the penalties currently imposed on families with two earners who have unequal earnings. For similar reasons, language of “concession”, “give-away” and so on is inappropriate in considering such proposals. Any costs are costs that are borne by families suffering under the current regime.

Of course, a change in the tax system would reduce revenues to the Treasury. It is tempting for proponents and opponents of change to try to calculate precisely those revenue changes. This is unrealistic. Given that part of our case is based on the disincentives for family formation and the long-term costs for society as a result, it is simply impossible to assess the behavioral changes and likely impact on both public spending and taxation over a generation or two. As such, we should be very careful not to put too much focus on estimates of first-round effects. It is almost certain that the long-term reductions in revenue to the Exchequer will be lower than the short-term reductions in revenue. There will also be consequent reductions in government spending from our proposals.

Even in the short term, the revenue losses of change may be less than expected. Taxing family rather than individual income will reduce, or remove entirely, the incentive for families to organise their affairs in tax efficient ways such as accumulating pension savings or making charitable donations only in the name of the main earner (normally the man) whilst other savings are built up in the name of the non-earner. Indeed, in this respect, our current individualised tax system, paradoxically, discriminates against women by reducing their autonomy (in the case of charitable gifts) and opportunities to build up pension savings.

The revenue losses of a complete change in the tax system so that it mirrors the German system would be impossible to calculate using published data or models. We can, however, make some estimate of the cost of allowing the whole of the nil-rate tax band to be transferable. The practical effect of such a change would be as follows:

  • The household could elect to be taxed on the basis of family income
  • The tax-free allowances would be combined (to give a tax-free allowance of around £25,000 for the sake of illustration)
  • £37,500 would then be taxed at the basic rate of income tax (the basic rate bands would not be combined if only the nil-rate bands were combined)
  • Higher rates of tax would apply above a family income of £62,500
  • Combined family charitable deductions and pension contributions would apply to the family income
  • All National Insurance contributions would still be based on individual incomes, though the disconnection between contribution rates and benefit accrual also leads the National Insurance system to discriminate against families with an uneven split of earnings between the adults.

It is essential that there is no “clawing back” of the transferability of the tax-free allowance when higher incomes are reached as currently happens with the small marriage allowance. This would undermine the integrity of the proposed system and create unacceptably high marginal rates of tax at what are, in some parts of the country, relatively moderate family incomes.

Various estimates of the revenue reduction from making tax-free allowances fully transferable (or, as we prefer to frame the approach – “moving tax assessment to be based on family income and ability to pay”) have been made.

The starting point for estimates is the so-called “cost” to the Treasury of the current Married Couples Allowance which is 10 per cent of the nil-rate band or personal allowance. According to Seely (2022), latest estimates of that are £560m a year.38Anthony Seely, (2022), “Income tax allowances for married couples”, House of Commons Research Briefing. (Accessed 30th Jul. 2022). Don Draper, from the Tax and the Family group suggests that fully transferable personal allowances would cost around £6 billion39Don Draper, Jul 21st 2022, “Truss plans tax cut for families”, /2022/7/21/truss-plans-tax-cut-for-families (Accessed 30th Jul. 2022). . However, this was based on the assumption that the current restrictions surrounding the transferability of the allowance were maintained. This includes its withdrawal for higher rate tax payers and it non-applicability to cohabitees. We propose that the whole tax system is moved to a household income basis regardless of whether the adults are married and regardless of whether there are dependants.

Even if, in practice, only the personal allowances were aggregated within the household, more people would benefit than currently benefit from the married couples allowance. In addition, it is an integral part of our proposal that other tax bands would be automatically pushed up (even if not aggregated across the household). This would mean that taxpayers would benefit from the aggregation of allowances at their highest marginal rate. These factors would raise revenue losses above £6 billion. On the other hand, there would be a number of mitigating factors, in addition to the long-term behavioural change discussed above:

  • The full benefit of the current marriage allowance can be obtained if the lower earner within a family is earning up to approximately £11,300. However, the full benefit of combining tax-free allowances will not be felt by a family if the lower earner is earning any sum however small.
  • The full benefit of the current marriage allowance can be obtained if the higher earner is earning over approximately £13,800. However, the full benefit of combining tax-free allowances will not be felt by a couple unless their combined earnings are above £25,140 after pension contributions, charitable donations etc.
  • Couples will have fewer incentives to make pension contributions, investments and charitable donations in ways designed to minimise family tax bills.
  • There would be lower costs of childcare subsidies provided to families and less political pressure on government to extend subsidised or free childcare down the age range.

This does, of course, lead to the question of how any change in the tax system should be financed. Political considerations tend to discourage governments from reforming the tax system in radical ways. We would suggest that the government is open and honest. For a given amount of revenue raised, our tax system taxes some families too heavily relative to others. This may mean some reduction in the personal allowance or increases in marginal rates in order to finance radical change. Such change would have long-run benefits, but it would also make the system fairer according to the principles we have set out. The net losers would be two-earner families where both members of a couple earned similar amounts of money and single people.

Next chapter

Chapter 8: Conclusions

A tax system has many different objectives. These objectives include the funding of goods and services that have to be provided collectively (or which governments have chosen to provide collectively) and the redistribution of income. It is generally accepted that taxes should be levied based on the ability to pay for both of these purposes. This is even true of most people who believe that taxes should be simple and flat – normally flat tax systems have a tax-free allowance so that those on lower income do not pay tax.

However, the principle of ability to pay is breached in a serious way in the UK tax system. Quite rightly, when we measure poverty and inequality, we look at family or household income. We do not say that an individual is poor if they happen to be a non-earner in a family that earns £1m a year. Despite this, when we levy taxes, we do not levy them on the basis of family income but on the basis of individual income. This principle, which is relatively recent in the UK, means that a family with two earners will pay much less tax than a family on the same income with one earner or where the two main adults earn very different salaries. Typically, such families will have one adult who is caring for children or for elderly relatives.

The extent to which we penalise families which have two adults on different incomes is very significant. A single-earner family on £30,000 a year would have to earn an additional £4,988 per annum to have the same disposable income as a dual-earner family with a 50-50 split of incomes. A single-earner family with a gross income of £70,000 a year would have to have gross earnings of 20% more than a dual earner family in which both adults have the same earnings if they are to have the same net earnings. These penalties rise further as household incomes rise. Low-income families are not hit as badly in cash terms, but the impact is surely greater. A single-earner family earning £25,000 per annum, for example, would pay nearly £2,500 in additional income tax than a two-earner family with the same household income. They would also pay higher national insurance contributions and yet only accrue one state pension. These differences between families on low incomes with different household earnings composition are huge given the limited spare disposable incomes that such families have.

People often see the opportunity to be a single-earner family as a luxury that can only be enjoyed by the rich. This has not always been the case. However, our tax system benefits most those families where there are two earners on high salaries. In other words, where the spouses both have successful careers. A situation in which one partner combines part-time work with caring responsibilities or where one partner has more limited career opportunities or aspirations seems to be penalised with alacrity. This seems to reflect a cultural bias amongst out political elite against work in the home yet reliable survey evidence suggests that over one-third of working mothers of pre-school children would prefer to be at home full time if they could afford it and two-thirds would work fewer hours outside the home40Childcare and Early Years Survey of Parents, 2019: It is worth repeating that what is proposed here does not provide some special benefit or concession for families where one parent works within the home – it would restore neutrality and treat all families the same no matter what the split of earnings between the couple is.41This reporting of Liz Truss’s proposals is a good example. The impression is being given that some kind of concession is being given to single-earner families or families with an uneven split of earnings:

The interaction of the tax and benefits system reveals an even greater range of problems – especially for those on low incomes. Because benefits are withdrawn as household income rises but taxes are levied based on individual earnings, if two parents of one or more children should marry or live together, they could well find that their income reduces. In other words, the government penalises family formation just as it might penalise smoking, alcohol or sugar consumption – as if it is a bad thing.

There is a solution to these problems. In many countries, tax-free allowances and tax bands for couples could be aggregated within a family. Additional tax-free allowances (at different levels) could be given in respect of children to replace welfare benefits for children. The scale of progressive tax rates could then be applied to the aggregate family income. This approach, similar to that which is adopted in France and Germany, would mean that all families would pay the same amount of tax whatever the split of earnings between the adults in the family. We believe that this would create a situation in which families were taxed fairly on the basis of ability to pay and not on the basis of their choices regarding how paid work was split between the main adults.

Next chapter


  • Apps, P., Rees, R. (2018) Optimal family taxation and income inequality. International Tax and Public Finance, Vol. 25, p. 1093–1128.
  • David Cameron, (14th January 2019) House of Commons Briefing Paper Number 4392: Tax, marriage & transferable allowances”, p. 32.
  • Don Draper, (2022), “Truss plans tax cut for families”, /2022/7/21/truss-plans-tax-cut-for-families (Accessed 30th 2022).
  • Friedrich Hayek, (1960) The Constitution of Liberty, Chicago: University of Chicago Press.
  • Friedrich Hayek, (1976) The mirage of social justice. Chicago: University of Chicago Press.
  •, (2021) Guide to French Income Tax. Available from: (Accessed 21st Nov 2020).
  • Geiger, BB; Scullion, L; Summers, K; Martin, P; Lawler, C; Edmiston, D; Gibbons, A; Ingold, I; Karagiannaki, E; Robertshaw, R and de Vries, R (2020), “Should social security reach further? A study of those not claiming benefits at the start of the COVID-19 pandemic”, Welfare at a (Social) Distance Project Report.
  • GOV.UK, (2021) “Marriage Allowance”, Available from: (Accessed 17th July 2021).
  • GOV.UK, (2021) Benefits calculators: Turn2us. Available from: (Accessed 8th Feb 2021).
  • Institute for Fiscal Studies, (2010) Press Release: Couple ‘penalties’ in tax and benefit system are widespread, but almost impossible to eliminate. Available from: (Accessed 8th Feb 2021).
  • International Federation for Family Development – IFFD, (2017) The Crucial Role of Families, Available from: (Accessed 3rd May 2021).
  • Jonathan Williams, (2019), Taxing Families in the UK, Institute for Fiscal Studies.
  • John D. Mueller (2014), Redeeming Economics – Rediscovering the Missing Element, ISI Books, Delaware, USA.
  • Kristian Niemeitz (2010), Transforming Welfare – incentives, localisation and non-discrimination, Discussion Paper Number 32, Institute of Economic Affairs, London.
  • Leonard Beighton, Don Draper and Alistair Pearson, (2018) The Taxation of Families, Care – Tax & the Family.
  • Meade, J. E. & Institute for Fiscal Studies (Great Britain).  (1978).  The Structure and reform of direct taxation: report.  London; Boston:  Allen and Unwin.
  • Mirrlees et al. (2011) Special Issue: Special Issue on the Mirrlees Review. The Journal of Applied Public Economics, Vol. 32(3), P. 319-453.
  • Musgrave, R. A. (1967) In Defense of an Income Concept. Harvard Law Review, Vol. 81(1), p. 44–62.
  • Nigel Lawson, (1992) The View from No.11: Memoirs of a Tory Radical, London: Bantam Press.
  • OECD, (2022) Income inequality (indicator). Available from: (Accessed 05 January 2022).
  • Office for National Statistics. (2020), Household income inequality, UK: Financial year ending 2020. Available from: (Accessed 14 June 2021).
  • Patricia Morgan, (2017) The War between the State and the Family: How Government Divides and Impoverishes, Abingdon: Taylor & Francis.
  • Pope John XXIII, (1963), Pacem In TerrisEncyclical Of Pope John XXIII on Establishing Universal Peace In Truth, Justice, Charity, And Liberty, April 11, 1963. Available from: (Accessed 7th June 2021).
  • Ruth Kelly, (13th March 2000) “Brown’s chance to take credit for reviving family life”, The Guardian. Available from: (Accessed 12th Nov 2020)
  • Sam Bromley, (26th June 2019), “Revised HMRC figures reveal more than 2 million people are missing out on this tax break”, Simply Business, Available from: (Accessed 23rd Nov 2020).
  • Antony Seely, (2022), Income tax allowances for married couples, Research Briefing, House of Commons Library.
  • Simon Alford, (2011), “Clegg attacks marriage tax break plans”, The Times, December 18th 2011.
  • Stuart Adam, Mike Brewer, (2010) Couple penalties and premiums in the UK tax and benefit system, Institute for Fiscal Studies, IFS Briefing Note BN102.
  • Tax & the Family YouTube Channel, (2018) An Interview with Lord Lawson about Independent Taxation. Available from: (Accessed 25th March 2021).
  • Tax and the Family, (n.d.) Professional reports that inform policy. Available from: (Accessed 5th April 2021).
  • The Holy Bible – New International Version (1979). London: Hodder & Stoughton.
  • Turn2us, (2021) Assessment Period. Available from: (Accessed 15th Feb 2021).
  • UK Parliament, (2021) Research Brief: Average earnings by age and region, House of Commons. Available from: (Accessed 28th Nov 2021).
  • Viktor Steiner, Katharina Wrohlich, (2004) Household Taxation, Income Splitting and Labour Supply Incentives – A Microsimulation Study for Germany, German Institute of Economic Research, DIW Discussion Papers No. 421.

Join our mailing list