Public policy and inequalities: lessons for policymakers from the IFS Deaton Review

Executive summary

The IFS Deaton Review has been a major interdisciplinary project studying the causes and consequences of inequalities. The Review has studied inequalities related to education, health, income, wealth, political power, geography, gender, ethnicity and social mobility. This report draws lessons for policymakers from the Review. In particular, we set out a high-level policy toolkit and apply this to four specific areas of public policy.

1. Introduction

How should policymakers design public policy when faced with different inequalities? Over the last six years, the IFS Deaton Review has brought together researchers across disciplines to study the causes and consequences of inequalities, with a particular focus on the United Kingdom. Seventy-nine papers and commentaries have been published and are available to all,1 and the lead authors of the Review have synthesised the findings of the Review into a new book (Johnson et al., 2026).

Building on this vast scholarly exercise, this report aims to help policymakers think about how to design good policies in the light of the different ways in which inequalities are generated and the trade-offs that exist in designing and implementing public policy.

There is – and always will be – reasonable disagreement over the extent to which policymakers should seek to reduce different forms of inequality. Where people want particular inequalities to be lower, they may have different reasons for that preference. Furthermore, many disagree on the extent to which they are willing to sacrifice other goals to achieve a reduction in inequality. Our aim is not to take a particular stance on when or to what extent inequalities should be addressed, but to help policymakers and others to understand and navigate the choices available.

A key thing to emphasise is that the IFS Deaton Review has not been a study only of income inequality. Instead, the Review has cast its net widely, examining inequalities in education, health, income, wealth and political voice/power, additionally considering between-group inequalities (such as those between men and women, and people of different ethnicities), geographical inequalities and social mobility. This list is not exhaustive.

The breadth here is vital because some people only care about some inequalities and not others. Johnson et al. (2026) point out that there have been inequality-reducing social and political changes, such as reduced tolerance for discrimination and improved property and voting rights for women, which now command widespread support. Concerns about the seriousness of some other inequalities are common but not quite so widespread. Benson et al. (2024) found that many people considered geographical and income and wealth inequalities to be more serious than health or generational inequalities. Even focusing only on one particular concept such as income inequality, concerns about the lower end of the income distribution being ‘left behind’ may be very different from concerns about the upper end of the distribution ‘racing away’. Inequality is not the same as poverty or deprivation, and many may be more concerned about reducing deprivation than they are about reducing inequalities more generally. How concerned people are about particular inequalities might also depend on how and why they arise: whether they arise because of effort, or luck, or discrimination, for example.

As we will discuss in the next section, which kinds of inequalities policymakers are concerned about becomes very important when considering public policy. But given the wide variety of inequalities and opinions about them, it is important to provide context for later discussion by setting out some key stylised facts about inequalities in the UK, which have been summarised in the previous work of the IFS Deaton Review:

• Income inequality is high compared with most high-income countries, though notably below the level of the United States (Bourquin, Brewer and Wernham, 2024). Income inequality rose substantially right across the income distribution during the 1980s, with much less change across most of the income distribution since then. The top of the income distribution continued to pull away until the late 2000s, since when top-end income inequality has not risen (Delestre et al., 2024).
• Standard measures of wealth inequality have been flat – at historically low levels – since the early 1980s (Bourquin, Brewer and Wernham, 2024). However, wealth has become much more important relative to incomes over time. Between the early 1990s and the late 2010s, household wealth almost doubled relative to GDP. Wealth accumulation is arguably harder for younger generations than it previously was for older generations. For example, people who owned property before the very large increases in property prices from the mid 1990s to 2007 have generally accumulated significant housing wealth. In contrast, generations reaching adulthood since then have seen much lower rates of homeownership and are often increasingly reliant on financial help from parents to buy a home. Inheritances are becoming an increasingly large share of people’s lifetime income.
• Wage inequality rose between the late 1970s and early 2010s, though it has fallen back a little following large minimum wage increases starting in 2016 (Blundell et al., 2023). Wage inequality is higher at older ages, with less educated people and lower earners (especially women) experiencing less wage progression as they get older. Lower earners and less educated people also have fewer training opportunities and fewer other workplace benefits and often face insecure work, with large numbers working in low-productivity solo self-employment. However, rising employment and lower household worklessness (at least up until the COVID-19 pandemic) meant many more households benefited from some income from work, pushing up incomes towards the bottom of the distribution.
• Economic inequalities mean that some children grow up in families with much less financial capacity to invest in their children’s education than others. These inequalities could also shape the wider resources that families have – for example, affecting parents’ time, stress levels, information or aspirations. While the state education system plays a very important role reducing the impact of inequalities between different families, there are still large and persistent differences in educational achievement between children from richer and poorer households (Farquharson, McNally and Tahir, 2024). However, there have been significant improvements in school performance in England in recent years, and very few countries have both better average performance and more equal performance for more and less deprived children. Average performance is lower in the rest of the UK and – in Scotland – less equal as well. The UK is also unusual amongst European countries that first- and second-generation migrant children do very well at school.
• There are significant health and disability inequalities in the population (Banks and Berkman, 2024). People who live in more deprived areas or are less educated are more likely to be overweight or obese, experience poor mental health, and die from ‘avoidable’ conditions. Life expectancy is notably lower in poor areas. These health inequalities partly reflect economic inequalities over the life course. But even to the extent that these are ‘only correlations’ (rather than causal relationships), they still reflect the fact that many poorer people experience poorer health on average as well as lower income and consumption.
• Regional inequalities – about which public concern is particularly high – are substantial and have grown significantly over the last 50 years (Benson et al., 2024). There are wide differences across the country in many dimensions: health and disability, income, levels of education and the amount of innovation that takes place. In a reversal from 25 years ago, young children growing up in London now do better in school than those growing up in the rest of the country. Regions outside London and the South East have low innovation and many low-productivity firms. High housing costs in the most economically productive areas are an important reason why average material living standards are less unequal across regions than productivity and earnings are, but they discourage some people (particularly those from poorer backgrounds) from moving to more productive places. These high housing costs also lead to high rates of child poverty in London compared with other parts of the country.

This is only a small selection of the findings of the Review. But even in this short summary there are important nuances. There is often a complex situation in which some elements of a particular inequality look different from other elements. Take the last example, of regional inequalities – London is both the richest and most economically productive region of the UK and one of the regions with the highest income poverty rates once housing costs are accounted for.

The rest of this report proceeds as follows. In Section 2, we discuss key considerations for policymakers who are seeking to develop policies to address an inequality that they are concerned about. This provides a high-level toolkit which should be useful for policymakers who place more or less weight on different inequalities. It draws on insights from the IFS Deaton Review as well as a larger body of work undertaken at IFS over many decades. In Section 3, we apply these considerations to four specific areas of public policy: educational inequalities, policy around ‘good jobs’, geographical inequalities and wealth inequalities. Section 4 provides a brief conclusion.

2. Inequalities and economic efficiency: a policy toolkit

When and how public policy should appropriately address inequalities is obviously a difficult question. This is not least because there are often underlying changes in inequalities that are not directly related to public policy; there is not agreement about exactly which inequalities the government should seek to reduce; there are complex interactions between different inequalities; and there will sometimes be trade-offs between different types of inequality, and between reducing inequalities and other objectives. One critique made by Johnson et al. (2026) is that some past policy decisions have not paid sufficient attention to the effects on inequalities.

The broader lesson is that policies almost always have a range of effects that need to be considered. Policies aimed at, for example, boosting economic growth should be considered in light of their expected effect on inequalities and other aspects of fairness. And policies designed to reduce inequalities should not be made without heed to their broader effects, notably including effects on economic efficiency. In this report, we discuss the relationship between inequalities and economic efficiency and how good public policy can be made given these relationships.

It is important to be clear about what economic efficiency is and is not. An efficient outcome is one in which resources (such as people’s time and effort and natural, physical and financial capital) are being used in such a way that it is impossible to rearrange production or consumption to make someone better off without making someone else worse off.2 That an outcome is efficient does not imply that it is necessarily fair or desirable, merely that there are no easy, mutually beneficial changes left to make. Where there are economic inefficiencies in an economy, it implies that – at least in principle – it would be possible to make some people better off without making anyone worse off. A particular policy would be described as improving efficiency if it creates enough total gains that the winners could, in principle, compensate the losers and still be better off – even if that compensation does not actually happen in practice.3

There are two key reasons why efficiency considerations are relevant to discussions of inequality.

First, it has long been known that there are many situations in which there are ‘equity–efficiency trade-offs’: increasing some dimension of equity can come at the cost of reducing efficiency. A classic example is of income-related taxes and benefits: such redistributive policies can reduce some forms of inequality but can also, for example, lead people to work less, invest less in capital or skills (human capital) or choose less productive jobs or investments. The resulting inefficiency can sometimes show up as reduced national income. The net effect on society’s overall wellbeing can therefore be ambiguous, reflecting a trade-off between the gains from greater redistribution and the costs of increased inefficiency. In addition to this, Johnson et al. (2026) argue that lower economic growth can itself increase some forms of inequality. This is because weaker growth may increase the importance of inheritances or other financial transfers from parents (thereby reducing social mobility) and because weaker growth can make it harder to raise the taxes to fund inequality-reducing parts of the welfare state. As a result, there may be cases where reduced inequality today comes at the expense of higher inequalities in the future through hampering economic efficiency.

Second, and as Johnson et al. (2026) also highlight, many forms of inequality are directly caused or exacerbated by ‘market failures’ – i.e. there are problems that can cause economic inefficiencies and inequalities. In such cases, addressing economic inefficiency may be one of the best ways to address inequality.

Market failures will be a theme that we return to throughout this report. Broadly, a market failure occurs when private markets, left to themselves, lead to an inefficient allocation of resources. Key examples include externalities, where individuals or firms impose costs or benefits on others that are not reflected in prices (such as pollution); public goods, which are necessarily available to everyone and tend to be under-provided (or not provided at all) by markets (such as national defence); market power, where, for example, firms can restrict output and raise prices above (marginal) cost; imperfect information (including information asymmetries, where one party to a transaction has better information than another, such as in financial markets); and incomplete markets, where some services, or insurance against some risks, are not provided at all.

There can also be coordination failures in markets. These occur when individuals’ and firms’ actions are individually rational but collectively inefficient, such that the economy becomes stuck in a low-wellbeing equilibrium even though a better outcome is possible. Prices alone may not be enough to help people and firms coordinate their decisions. Coordination failures are more common when success depends on many people or firms making similar choices at the same time or when the best choice for one person or firm depends on what others choose. To give a simple example, workers and firms may each be willing – in principle – to move to a new area, but only if the other does so first; existing wages and prices may not provide enough incentive for either to move first.

A related set of issues concerns insights from behavioural economics. If individuals sometimes make decisions that leave them worse off by their own assessment – perhaps because of inertia, or not taking account of all available information, or overweighting present costs and benefits at the expense of the future – markets may fail to deliver an efficient allocation of resources. The kinds of arguments this report makes are relevant for designing policy in these settings too, though in some cases they make it more difficult to define what an efficient outcome in fact is (e.g. see Goldin and Reck (2022)).

It is important to note that market failures are not the same as outcomes that simply include some ‘bad’ aspect – it is only a market failure if an alternative allocation of resources could in principle be arranged which would improve some individuals’ situations while making no individuals’ situations worse.

In practice, market failures are common and in such cases there is a potential role for government intervention. Ways in which governments can address market failures and improve efficiency include providing effective institutions which underpin market transactions, such as a well-functioning legal system to enforce contracts and protect property rights; directly providing ‘public goods’; providing insurance or credit where the market would not due to information asymmetries; addressing the existence of positive or negative externalities; encouraging market competition or regulating uncompetitive markets; or ensuring provision of information to facilitate good decision-making.

The rest of this section provides a simple toolkit to help policymakers navigate policy choices where there are both inequality and efficiency effects.

Toolkit: guiding questions

The set of potential policies available to governments is vast. At one end of the spectrum it includes measures that seek to establish and/or shape private markets, including labour, capital, housing, insurance and other product markets. Examples of specific policy tools here include the protection of property rights, regulations – including price controls such as minimum wages – and competition policy. At the other end of the spectrum are policies in which the state directly finances or provides goods and services, such as healthcare and transport infrastructure. These can be either free at the point of use (such as state education or the NHS) or subsidised (such as social housing). The government also operates a large tax and benefit system. Some tax policies are explicitly intended to shape market behaviour – most notably, a growing list of ‘sin taxes’ can be motivated by a desire to address externalities. More broadly, raising revenue through the (progressive) tax system and using it to fund both welfare payments and public services is a key way that the government redistributes resources.4

Broadly speaking – and as a way to navigate an otherwise unwieldy set of considerations and trade-offs – we suggest three questions that policymakers who are considering how to address inequalities they are concerned about can ask themselves to help identify which policy tools may be most effective. The three key questions are:

1. What is the source of inequality and does it involve a market failure?
2. What are the interrelations between, and effects of, inequalities?
3. What is the least inefficient way to address the inequality, and are the costs of doing so exceeded by the government’s concern about this dimension of inequality?

We examine each of these questions in turn.

1. What is the source of inequality and does it involve a market failure?

Understanding the sources of inequality is an important first step to considering policy solutions. When an inequality is being driven or exacerbated by a market failure, the appropriate policy response – subject to being able to design an effective policy in practice (more on this below) – will usually involve directly addressing the failure, rather than allowing the failure to continue and trying to use other policies, such as taxes and benefits, to reduce income inequality (or another form of inequality). This will require understanding exactly what the failure is – that is, precisely why markets are not producing efficient outcomes.5

For example, if policymakers are concerned about top income inequality, part of that inequality (perhaps the part that is more concerning to people) results not from high work effort but from market power that leads to high profits accruing to small numbers of high-income individuals. Significant market power (in extremis monopoly) also leads to inefficient allocation of resources because firms restrict output and raise prices compared with more competitive markets. Rather than using redistribution through distortionary taxation to address this, effective use of competition policy could improve efficiency and reduce top income inequality through lower profits accruing to small numbers of people.

As in this example, if there is a market failure that causes or increases a particular inequality, policy may be able to both increase economic efficiency and reduce an inequality. However, many inequalities are not a result of market failures. Instead they can arise because people have different abilities, access to resources, or preferences, leading them to make different choices. These are key factors in leading to unequal outcomes. Some policymakers will take a view that aspects of the resulting distribution of outcomes are inequitable, but in this situation policymakers will need to confront harder trade-offs between economic efficiency and desired distributional outcomes.

In other cases, inequalities may arise because of discrimination: for example, employers or customers might have discriminatory preferences against workers with particular demographic characteristics, resulting in such workers earning less than their counterparts.6 Policymakers might address this sort of earnings inequality ‘directly’, by regulating certain behaviours (e.g. making sex discrimination in pay illegal), or through the same kinds of redistributive policies as used to reduce inequality that arises through other means. Both of these will involve certain trade-offs. A rather different approach would involve measures aimed at shifting norms and attitudes, potentially without any negative effects on economic efficiency (e.g. Bursztyn, González and Yanagizawa-Drott, 2020).

2. What are the interrelations between, and effects of, inequalities?

There are many different inequalities which a government may be concerned about. The concern policymakers attach to different inequalities will vary depending on the preferences of the government. If one inequality rises as a result of a policy change and another falls, it will be important to consider which situation the government prefers. For example, policies that successfully increase productivity in cities in Northern England could reduce geographical inequalities but increase earnings inequalities. Or, as noted above, sacrificing economic efficiency today for more equal outcomes could lower economic growth and result in a greater reliance on inherited wealth for younger or future generations.

In other cases, rather than there being a trade-off between inequalities, they can be interrelated and mutually reinforcing. For example, Johnson et al. (2026) argue that ‘health inequalities … appear to derive from inequalities in education, earnings, wealth, and status’.

Where inequalities reinforce one another, they can combine to create traps of disadvantage. For example, a young person may be in a low-income family that is unable to provide educational resources; at the same time, they may live in a place where job opportunities are poor, such that the expected returns to education are low; at the same time, poor attainment may lead to poor health. In such cases, it may not be enough to tackle one problem in isolation; for example, alleviating credit constraints that prevent young people accessing education may be of little value if there continues to be little return to investment because job opportunities remain weak. In such cases, what may be required is a package of policy measures that can create a new equilibrium.

In this sense, there are existing inequalities (such as across families, or in different geographies) that, especially when combined together, can generate inefficiencies – the young person may end up being less productive through their working life, making them and society at large worse off. Such inefficiencies could be corrected through policies to reduce inequalities (such as redistribution to low-income families) or through policies more narrowly tailored to solving the market failures directly (such as providing educational resources to children from low-income backgrounds).

3. What is the least inefficient way to address the inequality, and are the costs of doing so exceeded by the government’s concern about this dimension of inequality?

As discussed above, where the inequalities of concern are caused by market failures, the government should aim to correct the market failures directly, at source if possible. Often this can both improve efficiency and reduce inequalities.

But when policymakers want to address an inequality that is not driven by a market failure (or by a suboptimal prior policy), there will likely be an equity–efficiency trade-off. This can arise directly when a policy that intervenes in a well-functioning market reduces efficiency. For example, in a competitive labour market, introducing a minimum wage can reduce wage inequality, but also reduces efficiency (by causing unemployment).

The trade-off can also arise indirectly: when policies are funded by distortionary taxation – taxation that changes (relative) prices, such as wages, and thereby changes people’s behaviour – that is a cost that needs to be taken into account. The direct distortionary costs of means-tested provision of benefits or public services are fairly well known: by limiting eligibility to people with lower income or wealth, policymakers will weaken incentives to earn and save. Provision of universal benefits or public services does not directly cause such distortionary effects. However, because spending on these policies is funded through distortionary taxation, the policymaker should be clear-sighted that any benefits from spending should be traded off against the efficiency cost of raising additional revenue through distortionary taxation. Therefore, any new spending plans need to be a sufficiently good use of public funds to justify reducing efficiency through greater taxation, and be a better use of public funds than other spending alternatives.

When there are multiple policy options that will achieve a similar distributional outcome, then naturally one ought to choose the most efficient. But often there will be some policies that do more to both reduce inequality and reduce efficiency, and others that do less. Here, policymakers have to take a view on where they want to be on the equity–efficiency trade-off – a political question rather than an economic one.

Even armed with a clear understanding of the source of an inequality (or inequalities) and the costs and benefits of using different policy tools, there will often still be a choice over which policy tool to use. There is not a one-size-fits-all answer to selecting policies; the case studies in the next section work through some specific examples. In all cases, there are two important practical considerations.

First, it may not be possible to design and implement a good policy and there can be risks of government failure. Governments are not perfect and there may be technical, informational or administrative constraints that make some policies unfeasible in practice.Governments face informational constraints which means they may not correctly assess costs and benefits. Public sector workers sometimes do not have the incentives to implement particular policies in the way intended. Implementing some new policies may be slow and costly (and frequently slower and costlier than first realised).

These all mean that while there may be a good rationale for a well-designed policy, the one that actually gets legislated may be far from well-designed. There are also risks that decisions are made for narrow political purposes. Some policies may be particularly at risk here, such as place-based policies, where political incentives to direct public-sector investment (or encourage private-sector investment) in specific areas of the country may be overwhelming. Governments may face significant lobbying from private firms and industry bodies, trade unions and third-sector bodies, which can create political pressure to take (or not take) a particular course of action rather than the best course of action.

Second, a rationale for a type of policy intervention does not necessarily imply a rationale for a specific policy. If politicians are considering a new policy or a change in an existing policy, they should consider the rationale for (and the costs and benefits of) the specific new policy or change they are considering, rather than the rationale for a generic policy of that type. For example, monopsony in the labour market can potentially provide an efficiency justification (as well as an equity justification) for a minimum wage. That does not mean that there is necessarily an efficiency justification for a higher minimum wage than we currently have – the minimum wage may already be at or above the efficient point. Another example (discussed more in the next section) is that while there may be a good efficiency rationale for some place-based policies, that does not mean that all specific place-based policies are a good idea.

3. Case studies: considering inequality and policy in four areas

We now turn to applying some of the lessons from the toolkit to four case studies. We select these examples as they highlight how different aspects of the toolkit can be used to think about policy, and because they connect to the chapters in Johnson et al. (2026). In addition, all four are the subject of ongoing public debate; recent years have witnessed much discussion of devolution and levelling up, employment rights and the minimum wage, childcare policy, and the taxation of wealth. Accordingly, we cover inequalities in education and families, ‘good jobs’, geography, and wealth.

Though we consider all these case studies somewhat separately, they are all interrelated with one another and with other inequalities. Importantly, they all connect to health inequalities – described by Johnson et al. (2026) as the ‘canary in the coal mine’ because they reflect the impact of other inequalities in society. The direction of causation here can go both ways – for example, inequalities in childhood experiences may result in inequalities in health later in life; and inequalities in health may affect the kinds of jobs people can get and hence labour market inequality. One important way that the government tries to correct health inequalities is, of course, through healthcare. But these other inequalities are also likely to play an important role, and a government concerned about health inequalities would do well to consider these broader factors throughout the life course. We discuss the connection to health in the case studies on education and families, ‘good jobs’ and geography.

Education and families

The experiences that people have during childhood and adolescence shape their trajectory throughout life. Inequalities by educational qualifications are large and open early: by age 3, fewer than 10% of children in the poorest quintile of households were in the top quintile of cognitive skills, compared with over 30% of their peers from the richest fifth of families (Cattan et al., 2024).

These inequalities persist over the life cycle: while 90% of graduates in their mid 20s to their early 50s are in paid work, around 20% of men (and 40% of women) in their 30s with GCSEs as their highest qualification are not in paid work. Among those who are employees, median earnings grow much faster for those with higher qualifications – and, by their early 40s, graduate men earn two-thirds more than men with only GCSEs or below (Farquharson, McNally and Tahir, 2024).

There is also strong causal evidence that education boosts people’s employment rates, earnings and wider outcomes. For example, when England and Wales raised their school-leaving age from 15 to 16 in 1972, young people who were affected saw a 7% increase in lifetime earnings (Buscha and Dickson, 2015). Young people who just miss out on achieving a ‘C’ grade in GCSE English are 4 percentage points less likely to continue in education than their peers who just scrape by (Machin, McNally and Ruiz-Valenzuela, 2020). A range of studies aim to estimate the returns to achieving qualifications after controlling for prior attainment, finding substantial benefits to lifetime earnings from better performance at GCSEs (Hayward, Hunt and Lord, 2014), from further education qualifications (Espinoza et al., 2020) and from higher education (Belfield et al., 2018). These returns vary a lot across different types of qualifications, subjects and institutions, as well as for different types of learners (Britton, Dearden and Waltmann, 2021) – but, on average, qualifications do raise lifetime earnings.

Education also shapes a range of wider outcomes, from better physical health (Davies et al., 2018) to higher life satisfaction (Oreopoulos and Salvanes, 2011) to lower criminality (Bell, Costa and Machin, 2022) and greater civic participation, social trust and engagement in community life (Briole et al., 2025). Higher levels of educational attainment even causally reduce adult mortality rates, highlighting the relationship between inequalities across the life cycle and health and mortality inequalities (Lleras-Muney, 2005; Clark and Royer, 2013).

Education is also an important part of wider discussions around self-efficacy – the belief in one’s own capacity to shape one’s own life – and its links to status and dignity. As Johnson et al. (2026) conclude in their second principle for policymaking, ‘people value the capacity to stand on their own two feet, … [the] feeling that they can shape their own destiny’. Education is one of the most direct ways to build that capacity.

Taken together, this means that education arguably plays an important role in shaping not just the inequalities in individual-level outcomes (such as earnings or health). It also contributes to wider concepts of social inequalities, including in terms of relational equality, voice and dignity (Johnson et al., 2026).

These inequalities by education level mean that policymakers motivated to address inequalities should consider where inequalities in educational attainment arise. In the UK, education levels overall have risen dramatically, shifting the shape of the overall distribution of education (Farquharson, McNally and Tahir, 2024). There is also a nuanced picture for how educational inequalities relate to other dimensions of inequality. There is a large gap in achievement between pupils from more and less socio-economically disadvantaged backgrounds, and this disadvantage gap has remained largely unchanged for at least 20 years (and increased during the COVID-19 pandemic). In other areas, though, educational inequalities have changed substantially: while children and young people in London today on average outperform their peers in the rest of the country, 30 years ago London had some of the worst educational results in England. Local authorities in inner London have seen particularly rapid improvements in performance (Blanden et al., 2015).

There has also been a rapid increase in the number of children with special educational needs and disabilities (SEND) in schools in England, with substantial variation in who gets support and large overlap with other forms of inequalities. Boys, children from poorer families, and children from White British and Black Caribbean backgrounds are more likely to receive the highest tier of support (Latimer, Sibieta and Snape, 2025). And educational inequalities across groups do not always mirror the patterns found elsewhere in society: some groups that do well in education nevertheless have on average worse outcomes in the labour market and more widely. Since the late 1980s, girls have substantially outperformed boys at school – and since the 1990s, women have been more likely than men to complete degrees, with the gap increasing each year since then (Farquharson, McNally and Tahir, 2024; Higher Education Statistics Agency, 2022). While children from ethnic minority backgrounds start out school behind their White British peers, young people from most minority ethnic backgrounds are much more likely to earn an A-level qualification by age 19 and to progress to higher education (Farquharson et al., 2024).

Diagnosing the problem

What is the source of inequality?

A core principle of the IFS Deaton Review is that it investigates inequalities (and the interplay between them), rather than just a single dimension of inequality. In education policy, there is a clear distinction between inequalities in children’s environments and inequalities in children’s underlying abilities. Children are born to families that differ greatly in the resources they have available – not only money and material goods such as access to secure housing or technology, but also things such as parents’ skills and information, mental health and stress, and networks and values. The structure of a child’s family – whether there is one parent or two, or whether the child is raised by someone else entirely – can differ. These factors very directly shape the environments that children are exposed to, from the home they live in to the relationships they grow up with.

At the same time, children have different underlying abilities. Often, these are clearest when it comes to inequalities in children’s healthiness when they come into the world: some babies are born healthy and thriving, while others are born with severe physical health conditions or profound learning disabilities. But this is also true of children’s propensity to develop in other ways, such as their cognitive skills or wellbeing. Cattan et al. (2024) review the evidence on genetic influences on development, finding that they account for a substantial share of the variation in many adult outcomes.7

These twin inheritances (resources and abilities) could be correlated – for example, if parents with higher abilities both earn more and pass on higher abilities to their children. They can also influence each other directly, with a child’s environment shaping the way their genes are expressed (Meaburn, 2025). But they are distinct concepts and are both important in shaping children’s outcomes.

Even if policymakers focus on the distribution of – and effects of – resources (rather than innate abilities), they must still define which resources are in scope. Relative to labour markets, children’s environments feature both much heavier and very limited government intervention. In countries (such as the UK) with a state school system, the state quite directly controls the quantity, quality, coverage and allocation of the vast majority of education for most children. While private systems do exist (most obviously in the form of private schools, but also tutoring and extracurricular activities), most children spend most of their ‘formal education’ time in state-provided settings.

At the same time, the other major environment that shapes children’s early experiences is the family. Here, the state generally has quite limited and indirect tools to shape children’s experiences. The government can (and does) intervene through the social services system in more extreme cases. But for most families most of the time, a policymaker’s tools are largely indirect: providing information and support, or increasing families’ financial resources.

So policymakers need to decide whether their aim is to deliver an equal education, or whether they want to go further and try to offset some of the inequalities in the distribution of children’s family resources – both financial and material resources, as well as wider resources such as mental health or information.

Are there market failures in play?

There are clear, well-established and well-evidenced market failures that mean that – in a pure free market – society would invest too little in education. Families with limited resources to invest in their children cannot easily borrow to fund the costs of education, since there is no obvious collateral to put up: quite simply, lenders cannot repossess a (better-educated, hopefully higher-earning) child down the line if parents do not make good on their repayments.

More fundamentally, there is the question of who is making decisions and who is benefiting (or not) from those choices. At least in the earliest years of life, parents and guardians make decisions on behalf of their child. A vast body of empirical evidence confirms that most parents take account of their child’s interests, in both the short run and the longer run, when they make these choices. But parents may also be weighing up other factors (including their own preferences). And they might be operating with limited information about what their child wants (or will want in future), where their child’s particular strengths could lie, and what the best combination of resources would be to make the most of those skills.

For policymakers concerned about the development of inequalities that occur in adult life, the education system can be an attractive area to intervene in. Because education often comes towards the beginning of life, policies that target children’s development have the potential to prevent later inequalities from opening up. To the extent that these early-life interventions result in lasting change, they can also be easier to justify on a value-for-money basis, as there is a much longer horizon for the investments to ‘pay off’. Indeed, studies that compare the value for money of different social policies often find that programmes targeted at children and young people offer the greatest chance of a high return on public money (e.g. Hendren and Sprung-Keyser, 2020).

But there is also the potential to have too much educational investment. Education is to some extent a positional good: families care not just about the level of skills and qualifications their child achieves, but also how they rank compared with others (an issue highlighted by Breen (2024)). This can lead to something of an educational ‘arms race’, as families invest more and more to try to preserve their child’s place in the education distribution.

These decisions can be individually rational, especially if the rewards to ‘making it’ are high. But collectively, families and young people can end up spending more time, effort and money than is justified by the actual improvement in skills and wellbeing that results, as every family tries to outcompete the others. This could be true for a small subset of families (e.g. families competing to tutor their child through the 11-plus grammar school exams) or more broadly. Policymakers sometimes feel it is necessary to intervene to shift away from this ‘bad equilibrium’ of competitive over-investment. For example, many East Asian governments, including China and South Korea, have become so concerned by the very long days young people spend in tutoring centres that they have introduced curbs such as restricted opening hours and additional regulation.

This complicates assessments of the equity–efficiency trade-offs. For example, providing a bright-but-disadvantaged student with additional resources and opportunities within the schooling system might level the playing field relative to a similarly bright student from a better-off family, with greater access to resources at home. But if the latter student’s family responds by increasing tutoring, paying for summer work experience or moving their child into private school, the effect on equity becomes much more muted – and society as a whole could end up spending too much time, effort and money on investments that are not especially productive themselves, even if the initial investment in a disadvantaged child was.

These risks are sharpened in the presence of greater inequalities in wider economic outcomes, for two reasons. First, a more unequal distribution of resources across families means that better-off families have more ability to (over-)invest in their children. Second, a more unequal set of outcomes in the wider economy raises the stakes: it matters more to parents whether their child will be able to secure a good education and a good, well-paid job (Doepke and Zilibotti, 2019).

Policy options

Policymakers frequently have clear goals to increase educational equality, in particular by improving attainment at the bottom end. For simplicity, in this section we focus on a policymaker seeking to narrow the socio-economic gap in educational attainment (as this is one of the most long-standing dimensions of educational inequality, and the one that most clearly mirrors inequalities elsewhere in economy and society). But the arguments here generalise to other types of inequalities in educational attainment.

In England (and indeed the rest of the UK), there is a long tradition of a state education system that – in many ways – operates as an equalising force. The vast majority of children attend formal schooling for a similar number of hours, where they learn a similar curriculum (at least within each nation of the UK) and are assessed with similar exams. The education system also goes further to compensate for inequalities in children’s home environments: funding is distributed progressively, with top-ups for children from economically disadvantaged backgrounds or with additional needs. A national funding formula largely (though not entirely) collects and then distributes resources centrally, rather than tying education funding to local economic circumstances or private donations (as in other countries, including the US). There is a wide range of programmes aimed at improving the quality of education that disadvantaged children have access to and targeting more equal outcomes.

By contrast, the resources that children have access to from their families are much more unequally distributed. This does not just refer to financial and material resources; as Cattan et al. (2024) conclude, differences in crucial environmental factors such as maternal mental health or home learning behaviours remain stubbornly high between richer and poorer households.

Broadly, this means that a policymaker who wants to generate a more equal distribution of educational attainment has three options:

1. increase the progressivity of the education system (‘state resources’);
2. improve children’s home environments (‘home resources’);
3. shift the balance of these two environments so that children spend more time in the (more equal) state education system.

There is a huge range of policies that could fit under each of these themes. We discuss a few active areas of policy in each.

Changing the progressivity of the education system

The state education system already acts to reduce many of the inequalities between children from different family backgrounds. Even so, there are aspects of the school system where educational inputs have become less progressive, or that favour children from better-off families. While schools with more disadvantaged pupil intakes still spend around 25% more per pupil than those with the least deprived intake, that premium has fallen from 35% in the mid 2010s (Sibieta, 2021). The 10% of state schools with the fewest pupils eligible for means-tested free school meals (FSM) are more than twice as likely to have ‘outstanding’ teaching quality as schools with the most disadvantaged intakes (Farquharson, McNally and Tahir, 2024). A school choice mechanism that relies heavily on proximity means that better-off families are better able to afford houses in desirable school catchment areas.

Because policymakers control many of the inputs in the education system, they often have levers to fairly directly change the allocation of resources. The ‘pupil premium’ (which offers top-up funding to schools with more disadvantaged pupils) is worth 16% less in real terms in 2025–26 than it was a decade before (Farquharson et al., 2026). Changing this, or other elements of the funding system, could directly shift financial resources between schools with different levels of disadvantage.

Policymakers could aim to change the distribution of teaching quality by building on programmes such as Teach First (which recruits high-performing graduates to teach in disadvantaged communities) (McLean and Worth, 2023). Targeted retention programmes offer financial incentives to teachers in key subjects with teacher shortages (such as STEM courses), or to those teaching in disadvantaged schools, though with mixed evidence of effectiveness (Worth and McLean, 2025). Other countries have been more directive – in France, for example, teachers are allocated to schools through a highly centralised system.

Alternatively, policymakers could change the rules governing access to specific schools. England has a well-used programme of school choice (Burgess, Greaves and Vignoles, 2019) where households are able to set out their preferences over schools. Parents from different social backgrounds largely value the same attributes of schools (academic quality, closeness to home, and similarity of the pupil body to their own child’s demographics) (Burgess et al., 2015). However, with distance to the school used as a common criterion to allocate pupils to over-subscribed schools (Burgess et al., 2023), the most popular schools end up with de facto catchment areas, pushing up local house prices (Gibbons, Machin and Silva, 2013) and making access related to a family’s ability to afford housing in that neighbourhood.

Other systems use alternative choice mechanisms such as lotteries to weaken the link between location and education – which can also affect wider outcomes such as social networks and community (Silva and Wessel, 2025). Modelling based on English data suggests that school choice systems that reserve a quota of spots for disadvantaged students could improve poorer students’ access to more preferred schools (though this would also see some pupils attend a different, less preferred school) (Burgess et al., 2025).

Importantly, in all of these examples, policymakers need to consider not just the first-order effects of changing the distribution of resources within the education system, but also the knock-on effects on ‘home’ resources. Because education can operate as something of an ‘arms race’, a policy that makes the education system more progressive can incentivise families with greater resources to put more time, energy and money into providing greater opportunities for their child outside the state school system (or even removing their child from the state school system entirely). This can complicate the equity–efficiency trade-offs of these interventions.

Improving children’s home environments

Another option for policymakers is to change the distribution of ‘home resources’ (including financial resources but also information, parental time, parental wellbeing, security and many other dimensions). Traditionally, this has been an area where the state has been more reluctant to intervene. However, there are a range of interventions that policymakers could consider.

First, the UK – like many other high-income countries – has programmes to direct financial resources to families with children, particularly those with low incomes. Child benefit, the child element of universal credit (and its predecessors), and policies such as the Scottish Best Start Grant or Sure Start maternity grants all fit into this category. These benefits can be paid out to everyone, based on family circumstances, or tied to specific behaviours (such as accessing education or healthcare). They can also be delivered as cash for families to spend as they wish, or restricted to specific types of goods (such as the UK’s ‘Healthy Start’ scheme subsidising healthy foods).

There is a large and mixed evidence base on the impacts of cash transfer programmes on child development.8 Overall, reviews of this literature suggest that – in high-income countries – cash transfers early in life likely have a positive, but quite small, effect on at least some aspects of children’s outcomes (Page, 2024). Of course, the costs of funding these transfers – which can be considerable and entail efficiency costs (as discussed in Section 2) – need to be set against these relatively small benefits.

A second approach to improving home environments is to provide parents with targeted, one-on-one information and support (which can also include support to access existing programmes). This can be delivered through home visiting programmes; the Family Nurse Partnership, which offers two-and-a-half years of home visiting to first-time teenaged mothers in some parts of England, increased the share of children achieving a good level of development at age 5 by 6 percentage points (Robling et al., 2022). A related third option is for policymakers to work with families in group settings. Holistic early years services, such as Head Start in the US and Sure Start in England, can significantly and persistently improve children’s outcomes and reduce inequalities (Carneiro et al., 2025a; Cattan et al., 2025).

Changing the balance between home and education environments

Because resources in the education system are distributed significantly more equally than home resources, policymakers can also affect inequalities by changing the relative importance of these two environments in children’s lives. Where the formal education system plays a larger role in children’s development, outcomes are likely to be more equal than if parents’ and families’ resources played a larger role.

Shifting the balance between home and education environments has been an active area of policy in England. Over the last 25 years, the government has increased spending on funded childcare hours eightfold, offering much more publicly-funded access to formal early education and childcare settings (Farquharson et al., 2026). In the schooling years, the expansion of before-school breakfast clubs increases the time children spend in a school environment. Partly as a result, they boost overall attainment and lower inequalities (Crawford et al., 2019). Programmes such as the National Tutoring Programme aim to offer state-funded access to resources that had traditionally been privately purchased. And most recently, debates on the right length of the school holidays (e.g. BBC, 2025) centre on the evidence that disadvantaged children can fall behind their peers from more affluent families during long periods away from school (though this has been most convincingly demonstrated in the US, where summer holidays are much longer than in England).

Of course, the impacts of these shifts depend not just on what is provided, but also on who takes it up. Where a new education programme is seen as valuable, take-up can be higher among more affluent families (who may find it easier to access a programme or face lower stigma from using it). For example, take-up of early education is often higher among better-off families even when entitlements are universal (Cornelissen et al., 2018). This can blunt the inequality-reducing effects of these programmes. Interventions to improve outreach and uptake, or explicit priority for a more targeted set of families, can instead increase take-up among the more disadvantaged (though may have unintended consequences for stigma or broad-based appeal).

The right balance between ‘state’ and ‘home’ environments is not just a question of inequalities – there are also questions of the state’s resources (and the resources of the wider workforce) and wider debates about the appropriate role of state and family in children’s lives.

Summing up

Even at the start of a child’s life, the education system does not act on a ‘blank slate’. The allocation of resources within the education system matters. But the allocation of resources across families – and how they use those resources – matters too. And so, as Johnson et al. (2026) conclude, ‘inequality of opportunity […] and inequality of outcome are inextricably linked’.

No education system can perfectly compensate for the many differences in resources – financial, relational, social and informational – that children experience at home. Views will differ on whether, or how much, it should try. Even in a society that somehow reached the (unattainable) benchmark of complete equality of opportunity, we would still have to grapple with questions about fairness in the distribution of outcomes – not just in education, but in earnings, wellbeing, health and many more domains. And even if the distribution of outcomes today is considered ‘fair’, it can still generate inequalities of opportunities for the next generation. This underlines one of the central messages of the IFS Deaton Review: there are multiple dimensions of inequality and they interact over the life cycle and across generations. We should expect policymakers to look at a range of policy levers and to face up to the trade-offs they entail when balancing different priorities and preferences.

Good jobs

In April 2025, 10% of employees in the UK earned less than £12.44 per hour, while 10% earned more than £37.16 per hour – a ratio of 3.0. Once one takes into account differences in hours worked (which are both voluntary and involuntary), the differences are considerably larger: 10% earned below £11,800 annualised and 10% more than £69,500 – a ratio of 5.9.9 Including self-employed workers – whose numbers have increased substantially since 2009 (Delestre et al., 2025) – would make the picture look more unequal still, since such workers tend to be found at the very bottom and very top of the earnings distribution (Giupponi and Xu, 2020).

Policy has been very focused on both these inequalities in pay and simply how many people have jobs. Governments of different stripes have made raising employment a priority, with the current government even having an explicit target of an 80% employment rate for working-age individuals.

A focus on inequalities in pay and in getting people into work is understandable, but labour market inequalities and issues do not nearly end there, because a job is far more than just a wage. Researchers have increasingly been turning their attention to the wider bundle of characteristics that make up a good job beyond high pay, including ‘participation, agency and learning’ (Institute for the Future of Work, 2025) and ‘career progression, access to non-pay benefits, workplace safety, adequate economic security, autonomy, investment in skill development, flexibility and work–life balance, and a sense of belonging to a larger enterprise’ (Johnson et al., 2026). Poor outcomes along these dimensions are not infrequent: for example, 28% of workers in the UK describe themselves as having little or no autonomy in their job, while 13% have an insecure contract or volatile hours (Cominetti et al., 2023). This has become more of a policy focus in the UK: the government recently passed the Employment Rights Act which, among other things, seeks to increase job quality by mandating higher levels of fringe benefits (e.g. sick leave) and job security.

These broader job quality factors are important to workers. In the US, Maestas et al. (2023) show that, on average, workers value the ability to set their own schedule, or to have autonomy in how they go about their job, at equivalent to 9% and 4% of their wages respectively. Moreover, empirically, job quality tends to be positively correlated with pay (Cominetti et al., 2023). A consequence of this is that the headline statistics on pay inequality above do not tell us the full story about inequality in total compensation, when we take the latter to include all the things that make a job valuable to the worker.10 Beyond the labour market, job quality is also correlated with inequalities in health (e.g. Chandola, Brunner and Marmot, 2006; Benach et al., 2014), with researchers hypothesising that job-related stress, insecurity and a lack of control over one’s own work can play a role in worsening health outcomes.

Diagnosing the problem

For the purposes of this case study, we restrict our attention to policies that are more narrowly labour market focused. But even here, in order to select the right policy for addressing labour market inequalities, we must first diagnose the problem – and (following Question 1 of the toolkit) identify whether it involves a market failure. In broad terms, one can think of three reasons why we might observe some workers having poor job outcomes:

1. The worker has low productivity: the amount of value they can produce per hour, even at a job well suited to their skills, is low. This does not in itself imply a market failure (though their low skills could be themselves caused by a market failure).
2. The employer has significant market power: workers have limited capability to move to other firms, giving their employer a greater degree of power over the terms of employment relative to that in a competitive market (and resulting in a market failure).
3. There is a lack of information or there are frictions in searching for, applying for and participating in work: workers are unaware of alternative opportunities, or finding and applying for them is costly in some way, or doing the job itself would be costly (perhaps because it is too far away and workers face mobility barriers). These factors can potentially result in a market failure. This can also cause employer market power; if workers cannot easily obtain other jobs, that gives their current employer greater bargaining power.

If these circumstances apply, workers will tend to get paid less. But, importantly, the jobs they can get will also be worse on broader quality dimensions, not just in their pay. For example, those who are more productive can command higher wages, but also higher-quality attributes on other margins – accounting for the positive relationship between pay and job quality. Similarly, employers can use market power to reduce wages (e.g. Benmelech, Bergman and Kim, 2022), but also to reduce the quality of jobs: Adams et al. (2023) show that greater employer market power in the UK results in a greater prevalence of contracts, including zero-hours contracts, that provide flexible hours with no guaranteed earnings.

So, when policymakers observe low-quality jobs, they need to determine what the fundamental driver is in order to select the right policy remedy. For example, if low-quality jobs just reflect workers’ having low productivity, then mandating higher job quality (such as by raising minimum legal fringe benefits) will tend to result in lower wages and possibly reduce efficiency (Manning, 2001; Gruber, 2008).

Moreover, as discussed in Section 2, a rationale for policy intervention does not necessarily imply a rationale for a specific policy. For example, a policymaker may want to tackle low levels of non-pay job amenities on the grounds that employer market power will result in under-provision. However, ‘fringe benefits’ in jobs – such as pension contributions, maternity leave and paid holiday – are in many respects simply another part of the overall remuneration package. Since these are generally subject to a lower tax and benefit withdrawal rate than pay (sometimes zero), this would tend to mean that there is an inefficiently high provision of them, especially for workers facing higher effective tax rates. The same basic argument applies more broadly, to things such as workplace safety or schedule flexibility, which are also not usually taxed. This issue does appear to be empirically relevant: Arold et al. (2025) show that tax increases cause jobs to shift towards having more (untaxed) workers’ rights, including non-financial rights (e.g. scheduling flexibility), and lower (taxed) pay. Whether in any given setting employer market power or tax is more important – and so whether non-pay job amenities are under- or over-provided – is ambiguous, and the policymaker needs to discern this before embarking on a strategy to directly increase job quality.

Policy options

There are many policy tools that one could draw upon to reduce inequality in access to good jobs. For example, policies to strengthen trade unions or collective bargaining could go some way to offsetting employer power; incorporating labour market power considerations in the context of competition policy could have the same effect; and public employment services can alleviate some frictions in matching workers to jobs. The government is now in the process of implementing the Employment Rights Act which, as discussed above, seeks to strengthen mandated benefits and employment protections, among other things. This reform can be seen as responding to inequalities in good jobs. Beyond those tools that might be narrowly considered as interventions in the labour market, a broader set of policies affect labour market outcomes –transport, health and housing policy are all likely to influence the kinds of jobs individuals can access. For space reasons, we limit our attention here to a handful of policies.

Minimum wages and means-tested in-work benefits

We begin with two policies that, in the UK in particular, have been very heavily relied upon to address labour market inequalities: minimum wages and means-tested in-work benefits (which today come primarily in the form of universal credit). Both of these are aimed at directly addressing inequalities in income, rather than necessarily addressing the underlying problems discussed above that can result in low-quality jobs. Partly because of this, these policies come with other undesirable consequences. Here we briefly review the effects of these different policies and the trade-offs they entail.

First, the two tools target different groups. In-work benefits are means-tested against total family income and take into account a family’s ‘needs’ such as rent and children. In contrast, the direct gains from minimum wage increases accrue to those with a low hourly wage irrespective of their other characteristics. For example, a worker earning £60,000 a year with a non-working spouse, two children and a typical level of rent would receive some universal credit – but would be far above the minimum wage. Conversely, a minimum-wage worker might be ineligible for universal credit if they live with a high-earning partner. At least insofar as partners share their income, in-work benefits are more closely targeted towards those with low living standards and so are a better tool for addressing inequalities in living standards (Giupponi et al., 2024) – though, in some cases, minimum wages will address intra-household inequalities.

Second, the two policy tools imply different effects on labour market efficiency. Minimum wages can reduce market failures when there is significant employer power, by limiting the extent to which employers can exploit that power to reduce employment and wages. They thus offer the opportunity for a reform that increases efficiency and equity (though even here, tackling the root problem of uncompetitive labour markets, if possible, would be better still). At the same time, in more competitive markets – or in less competitive ones, if the minimum wage is pushed too far – they can reduce efficiency by pricing some workers out of the market and causing unemployment.

In the UK, the existing evidence has not found large employment effects from the minimum wage so far – suggesting limited impacts on efficiency – though there must be a point at which further rises do start to cause significant job losses.11 In-work benefits tend to reduce distortions in a household’s choice to be in work or not (by reducing the effective tax rate on entering work and lowering the wedge between the cost to the employer of hiring the employee and the net additional income the employee receives). But they increase distortions in decisions about increasing one’s hours or about a second earner entering the workforce (because they raise the tax rate on an extra pound earned). Empirically, in-work benefits in the UK and the US have, on average, tended to raise employment with little effect on hours worked (Brewer and Hoynes, 2020), suggesting that, in themselves, they have been efficiency-enhancing. However, as emphasised in our toolkit, such benefits need to be paid for by distortionary taxes, so the overall impact on efficiency is not clear.

Third, a consequence of the increase in employment that in-work benefits have tended to produce could be to allow employers to reduce pay, undermining their efficacy as a method to address inequality. The available evidence here is limited but what there is suggests these effects could be quite significant (e.g. Rothstein, 2010; Gravoueille, 2025). This highlights a way that the minimum wage can act as a complement to in-work benefits, by capping the extent to which wage reductions offset increases in in-work support. That said, other parts of a job – the aspects beyond pay that form part of what makes it a ‘good’ one – can be negatively affected by higher minimum wages. For example, in the UK, zero-hours contracts have been shown to rise following an increase in the minimum wage (Datta, Giupponi and Machin, 2019; see Dube and Lindner (2024) for a broader review). Such effects imply reductions in efficiency, by distorting the characteristics of jobs provided away from what employers and employees would otherwise choose.

Fourth, both policies tend to weaken incentives for progression. Minimum wages compress the hourly wage distribution (Giupponi, Ray-Chaudhuri and Xu, 2024). In-work benefits raise effective marginal tax rates, as some of an extra pound earned is lost in the form of lower benefits. In both cases, this reduces the return to getting a promotion or (in the case of in-work benefits) working more hours. From the perspective of efficiency, this matters because it means workers are not fully rewarded for increases in their productivity. One complication comes in the context of training: if workers need to reduce their work hours to train, then high effective marginal tax rates mean they can do so with only a small loss to their income.12 Moreover, full-time work has potentially significant long-run implications since part-time work does not seem to generate much in the way of human capital that delivers career progression (Blundell et al., 2016). This fact reduces the efficiency case for subsidising part-time work while also strengthening the distributional case, since it implies that these low-paid part-time workers are more likely to be persistently poor.

It is not surprising that minimum wages and in-work benefits have been so popular among governments as a route to improving outcomes among low-income workers – they are relatively straightforward to implement and they deliver effects very quickly. At the same time, in part because they do not tackle the root issues causing poor-quality jobs, they can come with efficiency costs, and do little for – or may even worsen – non-pay aspects of jobs.

Training provision to improve productivity

We therefore review one policy that can go some way to targeting a root cause – training provision to improve productivity. The major way that the government currently intervenes in this area is through the formal education system, at least part of the goal of which is to improve people’s labour market productivity as adults (the previous case study discusses education explicitly). Beyond that, the government funds a range of other interventions targeted at adults, including ‘essential skills courses’ to improve things such as English and numeracy through the Adult Skills Fund; subsidies to cover the cost of training apprentices, where the apprentice works and learns at the same time; short vocational training courses (up to 16 weeks) called ‘skills bootcamps’ where participants learn specific skills and which are free to all adults; as well as more targeted interventions, including six-week training and work experience schemes for those unemployed and on benefits (sector-based work academies). While here we focus on training programmes for those who have finished compulsory education, an open question is when it would be more effective for the government to encourage acquisition of these sorts of skills during school – perhaps the case is especially strong for ‘generic’ skills that are used in a wide variety of jobs, though clearly adult retraining will be needed when technological advancements reshape the labour market.

A potential justification for government policies on training is that there may be under-provision relative to the efficient optimum. For employees, credit constraints create a barrier against paying for their own training. And employers are unlikely to provide sufficient training (except training very specific to their firm) because of a free-rider problem: a trained worker can go to another firm, allowing that second firm to capture the benefits of the training investment made by the first.

This makes a case for the government to intervene to increase training. At the same time, it is worth keeping in mind the point that the toolkit makes regarding the interrelation of inequalities (Question 2). If, for example, one needs to be healthy and well trained to get a good job, subsidising training alone in the presence of significant health inequalities may do little to alleviate labour market inequalities, unless it comes with interventions targeted at health too. Conversely, if adult training is a substitute for a good childhood education, training interventions might be especially effective for people who received a poor education.

One way to intervene in training is to provide a cash incentive: Blundell et al. (2021) find that a modest subsidy to training for mothers of young children (who have particularly loose labour market attachment) could pay for itself, thereby enhancing efficiency and reducing inequality. Another approach would be for the government to more directly provide the training itself. Of course, here the risks around government failure discussed in Section 2 are particularly relevant, as the government may lack the right information on what makes for a good training programme. Nonetheless there are encouraging signs: examining 49 studies of training programmes for the unemployed, Card, Kluve and Weber (2018) find significant increases in employment rates, including a rise of 6.7 percentage points two-plus years after the programme (compared with, for example, 1.1 percentage points for job-search assistance).

What sort of training should the government subsidise (or provide)? First, training that leads to vocational qualifications appears to be especially beneficial (Blundell, Dearden and Meghir, 1996), as it most clearly strengthens the worker’s opportunities outside the firm they work for, allowing them to move to a better job or negotiate up their existing one. Even outside the subsidisation or provision of such training, providing accreditation of qualifications is one way the government can help overcome information failures. Second, there is a case for sectoral training programmes, which work in tandem with local employers in certain industries to train potential workers with the skills those employers need, and focus on providing skills that are certifiable and transferable to other firms – again, enhancing workers’ broader opportunities.13 The scale of effects here can be large: Katz et al. (2022) study a sectoral training programme in the US aimed at low-wage workers which resulted in earnings gains of 12–34%. Third, social skills (such as working in a team, leadership and taking responsibility for outcomes) seem to be key to progression for low-skilled workers (Aghion et al., 2024); basic literacy and numeracy also have high returns (Vignoles, 2016; Machin, McNally and Ruiz-Valenzuela, 2020). Training programmes – or stronger school-based education – that enhance such skills could be particularly valuable. Ideally, the training subsidised or provided would also target skills that complement new technologies (such as AI) – though how easy it is for the government to identify the relevant skills and the training that would develop them remains an open question.

Summing up

Jobs not only represent the main source of income for most people, but can also be a route to meaning and skill development or, conversely, a site of stress, physical risk and a lack of autonomy. It is therefore not surprising that policymakers have seen jobs as a key domain to focus their efforts on when looking to address inequalities.

‘Direct’ approaches to improving job quality – such as minimum wages, in-work benefits and employment regulation – can all have a role to play. But they do not in general deal with the fundamental issues that drive poor jobs, and as a result can have unintended consequences. Policymakers concerned about these kinds of jobs should seek to identify the key underlying problem: low worker productivity, employer market power or frictions in job search (or some combination thereof). With that information in hand, they can design solutions that strike at the root cause of the issue.

Geographical inequalities

Regional inequalities in the UK are large by international standards (McCann, 2020) and span multiple dimensions including pay and productivity, education and skills, and health and life expectancy.14 Around half of the working-age population in London and Brighton have a degree, compared with 15–18% in Doncaster, Mansfield and Grimsby (Overman and Xu, 2024). Men born in Blackpool can expect to spend just 51 years of their lives in good health, while those born in Richmond, London can expect to be healthy for 69 years (Office for National Statistics, 2026). Labour productivity is around 42% higher in London than in England’s second-largest city, Manchester – a much bigger gap than in other advanced economies such as Germany, France and Spain (Brandily et al., 2022).

Many of these inequalities are rooted in the labour market, where outcomes diverged in the 1970s and 1980s and have remained persistent since. This divergence reflects Britain’s shift to a service economy. Industrial jobs that were concentrated in the North and Midlands disappeared, and new jobs in low-skilled service sectors emerged all over the country. However, new high-skilled jobs in tech, finance and professional services, which benefit strongly from agglomeration – the positive externalities of firms and workers clustering together – emerged largely in and around London (Xu, 2023).

Agglomeration can give rise to coordination failures, which means that market outcomes are not necessarily efficient. Linking to Question 1 of the toolkit, this means that some geographical inequalities could stem from market failures. For example, complementarities between firms and workers mean that, in many cases, no individual firm or worker has an incentive to ‘seed’ a second hub outside London, even if a second hub would increase social welfare. Productive firms may not move because local workers lack the skills they need; and local workers lack these skills because there are no job opportunities that require them. In this sense, some areas experience traps of disadvantage, discussed in Section 2.

Diagnosing the problem

Sources of agglomeration

The classic literature on agglomeration emphasises economic gains from workers and firms sharing resources, matching better and learning from one another when clustered together (Duranton and Puga, 2004). But agglomeration effects go far beyond worker–firm complementarities. Amenities that emerge in high-skilled cities, such as restaurants, museums and theatres, become a pull in their own right and make these places even more attractive to high-skilled people and the firms that employ them (Diamond, 2016). Venture capital firms grow around successful places (Chen et al., 2009), bringing access to finance, scale-up knowledge and networks to other firms in the area (Daams et al., 2025). Successful cities tend to have greater local government capacity, as they are able to collect higher business rates (partly retained by the council) and recruit from a wider talent pool. This makes them more able to design and enact growth-enhancing policies.

More broadly, we can think of the interrelations between different kinds of inequalities – discussed in Question 2 of the toolkit – as wider agglomeration effects. Children who grow up in successful cities are likely to see the opportunities that a good education unlocks, and thus be encouraged to invest in education; those growing up in places with few high-skilled jobs may rationally conclude that returns to education are low. The resulting divergence in education outcomes entrenches the divergence in economic outcomes, and so on. International evidence from resource booms shows that the local composition of jobs causally affects educational attainment (e.g. Kovalenko, 2023). In the UK, participation in higher education is higher in local authorities in London than in other local authorities with equivalent GCSE attainment (Farquharson et al., 2024).

The interrelation between economic outcomes and health is another potential source of agglomeration. A lack of labour market opportunities can lead to poor health, especially poor mental health (Browning and Heinesen, 2012; Pierce and Schott, 2020), which further dampens productivity. Poor health in an area also tends to limit local government capacity, as statutory spending on social care and special education needs and disability (SEND) crowds out other spending that could potentially improve local economic outcomes.15

These agglomeration effects have culminated over many years in a virtuous cycle of growth in London and its surrounding areas, and the opposite in other areas. Moving to another equilibrium requires coordination across a wide range of actors. This is because even if it were socially desirable for an entire cluster of economic activity to move to another region, it is not in the interest of any one firm or individual to do so, unsure whether others would follow suit.

Geographical inequality and efficiency

There are a number of reasons why the UK’s current economic geography – with productive activity overwhelmingly concentrated in London and surrounding areas in the South East of England – may not be efficient. People who move to London benefit local residents through agglomeration. However, they also impose negative externalities – increasing traffic and pollution, pushing up house prices, putting pressure on local services and so on. At some point, these ‘congestion externalities’ could outweigh the benefits of agglomeration, but people do not weigh up these social costs and benefits when deciding where to live. Indeed, London can feel very congested, with among the longest commute times and highest average rents among European cities (Eurofound, 2020; Eurostat, 2026).

Because people are not perfectly mobile across the country, the concentration of productive jobs in London means that people in other regions may not make full use of their skills, leading to an inefficient allocation of workers to jobs. For example, a computer science graduate from a top university is likely to be more productive working for a global technology firm than in an IT support role for a local employer. Graduate shares have increased fairly evenly across the UK over the past three decades, but outside London the number of jobs that require a degree has not kept pace. This has resulted in falling graduate wage premiums and a rise in graduates working in non-graduate jobs in other parts of the country (Stansbury, Turner and Balls, 2023; Xu, 2023).

The mismatch between workers and jobs is likely to be particularly stark for people from disadvantaged backgrounds, resulting in low social mobility. For young people in deprived parts of the country, moving up the career ladder often requires moving to a city with better opportunities, in particular London. But the wage gains from moving often emerge gradually over people’s careers, whereas moving costs and higher rents need to be borne up front. This means that – as with any investment – children who do not have wealthy parents can face credit constraints, in this case locking them out of better career pathways and preventing some from making the best use of their talents.16

Consistent with this, young people from poorer families in England are less likely to leave their home town by age 27, and those who leave are less likely to move to a big city, than otherwise-similar young people from richer families (Britton et al., 2021). Deprived places such as Bradford and Blackpool have some of the lowest levels of social mobility, measured by the difference in earnings between children from poorer and richer neighbourhoods (Carneiro et al., 2025b). Selective migration is likely to be one explanation for this: in these places, moving away is a key route to higher earnings, and children from more advantaged backgrounds are more able or willing to move.

Policy options

Ultimately, whether the current distribution of economic activity in the UK is inefficient – whether it is too concentrated in London – is a question to which there is no easy answer. This means that there is no consensus on whether policy should lean into London’s success and seek to expand it further, or actively try to even out economic activity across the country (‘level up’ in the parlance of the 2019–24 Conservative government). This tension is at the heart of many debates on public investment – for example, on how to allocate transport spending.

On the one hand, the fact that London is so much more productive than other parts of the country suggests that policies that expand access to London could be desirable. For example, improvements in transport links between London and other parts of the country would enable more people to commute to jobs in the capital, which would give more people access to high-paid jobs, strengthen agglomeration in London and improve the allocation of workers to firms across the country. The economic case for High Speed 2 was partly predicated on linking workers and firms in the North and Midlands to the economic cluster in London.

Other policies to expand access to London include planning reforms to reduce restrictions on developing land in and around London – for example, on the Green Belt. Moving costs could be lowered by reducing or abolishing stamp duty. Policies to increase housing supply and lower housing costs in London would both expand the city and make it more accessible to people from modest backgrounds, which would improve social mobility. More speculatively, it could be possible to target high-achieving young people from outside London directly – for example, by offering relocation grants or loans, similar to the ‘Moving to Opportunity’ project in the US, but targeted at moves across regions rather than neighbourhoods.

These policies could alleviate some of the negative consequences of substantial geographical inequality in labour market opportunities – improving the matching of workers to firms, increasing incentives to invest in education for (some) children outside London and increasing social mobility. But they would not reduce the level of geographical inequality.

If the current distribution of economic activity is indeed inefficient with too much concentration in London, policies that merely expand access to London would further entrench a suboptimal equilibrium. Furthermore, as discussed earlier, polling shows that the public considers inequalities between places to be the most serious form of inequality in Britain today. To the extent that people care about differences between places, as opposed to differences in opportunities facing people who grow up in different places, expanding access to London would not address these concerns.

If the current spatial equilibrium is inefficient, or if equity concerns with regards to these geographical inequalities are sufficiently pressing, the government could aim to solve the coordination problem and move to a more equal equilibrium. This would involve place-based policies that bring together the necessary complements – workers, firms, financial capital, infrastructure and so on – to kickstart positive cycles of agglomeration. Under this approach, transport investment should not focus on linking Northern cities to London, but rather on increasing density in a city region outside the South East. This would involve improving transport links within a city, between a city and its surrounding areas, and with other nearby cities. The government’s recent investments in Manchester’s intra-city transport (the Bee Network) and in the Northern Powerhouse Rail connecting cities between Liverpool and York speak to this approach.

‘Big push’ policies that aim to shift the spatial equilibrium are popular in East Asian countries (Moretti, 2024). In Western countries, a classic example is the Tennessee Valley Authority (TVA), which began in 1933 and consisted of large-scale infrastructure investments to industrialise what was then one of the most underdeveloped parts of the US. Federal transfers totalled around $27 billion in today’s prices between 1940 and 1958, the most intensive period of activity. The TVA generated large gains in manufacturing employment that continued to intensify well after transfers had lapsed – suggesting the policy had successfully kickstarted agglomeration – and its long-term benefits are estimated at roughly 140–210% of its costs (Kline and Moretti, 2014; Moretti, 2024).

The success of the TVA is specific to its design and context. The kinds of policies needed to seed manufacturing hubs in the 1950s are different from those needed to grow a high-skilled services cluster, where agglomeration externalities stem more from knowledge spillovers and dense labour markets than from shared infrastructure (Faggio, Silva and Strange, 2017). Nonetheless, recent case studies on ‘turnaround’ cities show that it is possible to revive ex-industrial heartlands through place-based policies. Examples include Lille in France, Dortmund in Germany, Bilbao in Spain and Pittsburgh in the US, which all suffered severely from deindustrialisation but achieved significant growth in the last few decades, driven by knowledge-intensive service sectors such as finance, professional services, education and technology (Frick et al., 2023).

These case studies point to several common themes. They include comprehensive strategies that pull multiple policy levers at once – addressing infrastructure, tax incentives, skill development, start-up support and so on, as well as urban regeneration policies to improve the quality of life. Economic strategies in these cities built on their existing strengths, often focusing on enhancing the scientific and innovation-led features of traditional sectors and leveraging local universities. ‘Significant, stable and long-term funding’ is identified as a prerequisite. Local and regional leadership, as well as engagement of a wide range of local actors in the private sector and civil society, are also identified as important (Frick et al., 2023). In the UK context, this points to targeting significant resources, over a sustained period, at a few cities outside the South East that have the potential to achieve self-sustaining cycles of growth.

However, ‘big push’ policies are extremely challenging, for the reasons discussed in Question 3 of the toolkit. Political incentives often push towards spreading resources thinly across places to signal fairness, as was the case with the Levelling Up Fund (which funded hundreds of projects to the tune of just £10–20 million per project), or targeting marginal constituencies for electoral gain. Funds can get reallocated as governments change. Even with the best of intentions, governments may lack the information or capacity needed to kickstart viable clusters. If investments are not large or comprehensive enough to trigger agglomeration, they risk significant deadweight (subsidising activity that would have happened anyway) and displacement (shifting activity that would have happened elsewhere). Standalone place-based policies such as enterprise zones often do not generate positive net effects for this reason (Neumark and Simpson, 2015).

For the most deprived places that are unlikely to ever sustain productive clusters, a starker equity–efficiency trade-off is at play. The government could still choose to channel resources to these areas for equity reasons, accepting that some degree of deadweight and displacement is inevitable, and that the interrelations between different types of inequalities may render policies less effective. For example, policies to improve educational quality in Blackpool may have limited success at boosting local incomes and employment given the lack of job opportunities available. Efforts to ‘seed’ a second hub outside London could help somewhat – we would expect a thriving Manchester to improve outcomes across the North West, just as London helps to lift up the wider South East. Ultimately, however, some degree of geographical inequality is inevitable. The current distribution of economic activity may be too skewed towards London, but a world with no geographical inequalities – and therefore no agglomeration – is neither achievable nor desirable.

Summing up

Productive economic activity in the UK is overwhelmingly concentrated in and around London. Agglomeration benefits mean that no individual or firm has an incentive to ‘seed’ a second hub, even if one were socially desirable. This means that the current spatial distribution could be inefficient as well as inequitable – though there is no consensus on whether this is the case.

If the government wished to move to a more equal equilibrium – whether for equity or efficiency reasons – it would need to bring together the necessary complements (workers, firms, financial capital and so on) to kickstart agglomeration outside the South East. Recent large-scale transport investments in Manchester, as well as the Chancellor’s narrative on focusing on the Northern Growth Corridor in her 2026 Mais lecture, point to the government adopting this approach. The challenge will be in delivering a large and comprehensive enough set of reforms to shift to a new equilibrium, and resisting the political temptation to spread resources more thinly, especially when returns could take decades to materialise.

Wealth inequalities

Inequalities in the distribution of wealth are a topic of considerable debate. There is disagreement about how best to measure wealth,17 about the forces that have driven changes in wealth over time18 and about how, if at all, policy should respond. When there is discussion about policy, there is often too little attention given to the exact policy goals. In this subsection, we summarise some of the concerns people have about wealth inequalities – highlighting that it is important to distinguish concerns about wealth per se from concerns about the incomes used to build wealth or the spending that it funds. We discuss taxes on wealth, which are among the more prominent policies considered by those seeking to reduce wealth inequalities.

Diagnosing the problem

Despite disagreements regarding measurement, certain broad facts about the nature of wealth inequality in the UK are clear. Over the course of the 20^th century, wealth inequality fellsubstantially. By one estimate, the share of UK wealth held by the richest 1% fell from 72% to 20%, while the share of wealth held by the richest 10% fell from 96% to 56% (from the World Inequality Database). Since the 1980s, those shares and the overall shape of the wealth distribution have remained fairly stable. Yet while the share of wealth held by those at the very top of the distribution has been broadly flat, levels of wealth have grown considerably relative to income: over the last 25 years, household wealth has almost doubled relative to GDP (from about three-and-a-half times as large to seven times as large; see Bourquin, Brewer and Wernham (2024)).To illustrate that: since the mid 1990s, real earnings have grown by only 37%, whereas real house prices have grown by 150%. The increase in the overall level of wealth has not generated a change in wealth inequality in the sense of the share held by the richest, but it has increased the absolute gaps between wealthier and poorer people, and has altered wealth inequalities across many different groups – in particular between generations, regions and holders of different asset classes.

The two biggest components of household wealth (on any conventional measure) are housing and private pensions. Across most of the wealth distribution, these two components represent close to 90% of households’ wealth. At each end of the wealth distribution, the picture is slightly different. At the bottom, some households have no wealth at all (indeed the poorest 10% have negative wealth on average, i.e. their debts are larger than their assets) and, for those that do have positive wealth, a greater fraction is made up of physical wealth (e.g. cars, furniture and appliances), with very little (if any) housing wealth. In the top 10% of the wealth distribution, financial wealth (e.g. bank accounts and shares) is more important than in other parts of the distribution, while at the very top of the distribution (in particular those with more than £5 million of wealth) business assets are very important (Advani, Bangham and Leslie, 2021).

While there is broad consensus on the extent of wealth inequality in the UK, views differ significantly as to the degree to which such inequality represents a problem that policymakers should seek to solve. To clarify the issues, it is helpful to consider what drives differences in wealth across the population.

A large part of the variation in wealth across the population at a point in time reflects age.19 For most people, wealth is built up during working life, peaks around retirement age and is then run down during retirement and/or passed on to others. As a result, even if everybody had an identical life course, receiving the same amount and spending the same amount at each age, wealth would look quite unequal across the population as a whole simply because some people are in their 20s with little savings while others are in their 60s having built up a pension pot and owning a home outright (or nearly outright). Even at a given age, wealth can vary among people with the same lifetime income. Some people earn (or inherit) more of their money later in life than others; some people choose to spend their money later in their lives than others (or bequeath it) – that is, to save more. Differences in the gap between when money is received and when it is disposed of create inequalities in wealth at any given age between people whose lifetime incomes are the same. Policymakers may conclude that wealth differences of these kinds – driven by age and by the timing of income and spending – are benign.

Of course, wealth inequality also reflects differences in people’s lifetime incomes, which in turn can reflect differences in abilities, preferences, inheritances, expectations, opportunities, effort and luck, among other things. Inequalities arising from some of these things might be considered by different people to be more problematic than others.

Those concerned about wealth inequalities should therefore first clarify exactly what they are concerned about. Is all wealth inequality problematic – even differences at a point in time that just reflect people being at different stages in their life cycle? Is wealth accumulation by saving problematic and to be discouraged? Or is the concern confined to inequality of lifetime resources, rather than wealth per se? Is it overall inequality of lifetime resources that is of concern, or is the concern particular sources, or uses of, wealth? And is inequality across the whole distribution problematic – or is there particular concern about wealth at the very top of the distribution, or (net) debt or lack of wealth at the very bottom?

At a basic level, people might be concerned about inequality of financial resources simply because £1 in the hands of a poor person is more valuable than £1 in the hands of a rich person: the population’s overall wellbeing might therefore be higher if resources were more equally distributed. As Johnson et al. (2026) highlight, some have argued that wealth inequality can also be damaging to wider society – for example, that it can lead to undesirable inequalities in social status and relationships or that those with the very highest levels of wealth may be able to exert undue political influence. In all these cases, it is worth considering whether it is wealth per se that is the real source of concern or differences in income and consumption, and whether the appropriate remedy to such concerns is altering the distribution of wealth as opposed to taking action either to alter the distribution of income or to constrain the activities of the wealthy that are deemed undesirable – for example, by regulating political donations.

For some, the degree to which inequalities in wealth or lifetime income are judged undesirable depends on the reasons they arise. Wealth that is the product of inheritance or good luck may be seen as more or less socially desirable than wealth stemming from hard work, artistic talent or entrepreneurial ingenuity. For example, some people are particularly concerned about inherited wealth and its implications for equality of opportunity and the intergenerational transmission of inequality. Inheritances are forecast to account for an increasing share of lifetime income in the coming decades as a result of higher levels of wealth amongst older generations (Bourquin, Joyce and Sturrock, 2021). If policymakers are concerned specifically about inheritances and intergenerational transmission, rather than wealth inequality per se, the most direct response is to target policy directly at intergenerational transfers rather than at wealth. Inheritance tax is the most obvious tool for this, but many other policies already seek to narrow inequalities of opportunity associated with parental resources – state-funded schooling being an obvious example, as noted above. By the same token, policymakers might be less concerned if a billionaire business owner has become wealthy by serving customers’ needs more successfully than competitors than if they derived their income from exploiting monopoly power where there are barriers to competitors entering the market. Indeed, to the extent that inequalities of income and wealth stem from market failures, there is a good case for intervention irrespective of any desire to redistribute wealth. De Loecker, Obermeier and Van Reenen (2024) and Tirole (2024), for example, both highlight the role for competition policy to ensure well-functioning markets – potentially both increasing the size of the economic pie and leading to its being shared more equally.

Policy options

Influencing inequalities in wealth is a complex business. As highlighted above, there are different reasons that a policymaker may care about wealth inequality and some cases in which the concerns are actually about inequality of income or consumption – or about particular sources or uses of wealth – rather than about wealth inequality per se. Correspondingly, there are many possible objectives that policymakers might seek to achieve and a crucial first step is to ensure that these objectives are clearly stated. Which policy tools are best suited to addressing concerns will depend on whether, for instance, the intended goal is to reduce wealth inequality in general, or only reduce relative wealth at the very top of the distribution, or only redistribute when wealth was gained through certain means, such as inheritance.

The full arsenal of tools at policymakers’ disposal is far too broad to provide a full account of here. Almost any policy that alters flows of income will result in changes to the distribution of wealth. The tax and benefit system affects how inequalities in income translate into inequalities of wealth and consumption in a mechanical sense, and also affects the distribution of resources via its effects on incentives and economic behaviour. There are many different features of the tax and benefit system – from the rates and thresholds of income tax to the tax treatment of inheritance, capital gains and business profits to the taxation of different consumption goods – that can be adjusted to fine-tune its effects and trade off different considerations; the design of the system, not just the overall level of taxation, matters enormously, as discussed in the Mirrlees Review of taxation (Mirrlees et al., 2011).

But, as Johnson et al. (2026) emphasise, tax and benefit policy is only one part of the story.The provision of public services such as education, health and policing directly confers some of the consumption benefits that private wealth would otherwise be needed to enjoy while potentially improving individuals’ opportunities to earn higher incomes. Minimum wages and other labour market policies affect the availability and pay of different kinds of work. Infrastructure spending alters the geographical distribution of opportunities, in turn affecting wealth gaps between different parts of the country. Planning regulations influence the cost of housing and alter the wealth of existing property owners and the owners of undeveloped land. Competition policy may prevent individuals from capturing economic rents as a result of market power. Many of these issues, and others, were discussed in the earlier subsections on education, geography and good jobs.

Most of these policy tools are aimed at altering flows of income and spending in the economy, which in turn affect the distribution of wealth. To the extent that policymakers’ underlying concern is with the distribution of lifetime income and consumption – or, even more so, particular sources or uses of wealth – then tools targeted at these flows are likely to be best suited.

For policymakers concerned with wealth inequality per se (that is, as distinct from lifetime income inequality), there are policies that directly target individuals’ stocks of wealth rather than flows of income and expenditure. Concerns about promoting wealth provision for those who might otherwise lack it have motivated initiatives ranging from ‘asset-based welfare’ policies (Emmerson and Wakefield, 2001) to automatic enrolment into private pension schemes. Child trust funds were one example of ‘asset-based welfare’ in the UK which, between 2005 and 2011, involved the government making lump-sum contributions to savings accounts for newborn babies – with larger amounts for low-income families – which could only be accessed once the child turned 18. Automatic enrolment into private pension schemes was introduced from 2012 in the UK and widened pension participation particularly among lower-earning groups (Cribb and Emmerson, 2020).

Annual or one-off wealth taxes

The most widely discussed policy option for redistributing between individuals on the basis of their wealth is an annual wealth tax.20 Almost by definition, a progressive annual tax on wealth would reduce the share of wealth held by those at the top of the distribution.21 And even a tax on wealth charged at a flat rate would reduce the importance of wealth relative to income and the absolute size of gaps in wealth across the distribution (depending on how the revenue was used). However, there are practical and principled downsides to using an annual wealth tax.

As noted in our toolkit in Section 2, whether a government wants a policy in principle is not all that matters; it also matters whether it is possible to design and implement it in a well-functioning way. The most formidable practical obstacle to enacting a wealth tax is the difficulty of defining and valuing taxable wealth. Valuing wealth each year would be a huge undertaking for the government. It would be extremely difficult to value certain kinds of assets, such as private businesses (which are a large share of wealth for the very wealthy) and defined benefit pension rights. Daly, Hughson and Loutzenhiser (2021) discuss practical valuation issues in detail.

But it is important to note that this is not a problem that can be circumnavigated (at least fully) by administrative ingenuity and resources: it is partly a conceptual issue. There is no right answer to the question of when the expectation of future income should be counted as an asset today. Many private businesses derive much of their value from the work the proprietor can be expected to do in future; it is not clear whether that value should be included or stripped out when valuing the businesses, but neither would be straightforward. Likewise, is a social media influencer who earns money from content because of their large following simply earning income or are they exploiting a (taxable) asset? One could ask similar questions about everything from celebrities’ image rights to future music sales. Clearly, it would be unfeasible to tax everyone on their expected future earnings. But doing so only when those future earnings are crystallised into a legally identifiable entity would distort commercial decisions and would be unfair on those whose activity happens to take that form. Wherever the boundary was drawn between what did and did not constitute taxable wealth, any feasible annual wealth tax would inevitably create complexity, distortions, and unfairness around the boundary. These problems would be especially acute at the very top of the distribution, where wealth is less likely to take the form of arm’s-length assets that can readily be identified and valued – though of course the aggregate administrative and compliance costs of a tax would be lower if it applied only to a very small group of taxpayers.

Aside from concerns about defining and valuing taxable wealth, an annual tax on the stock of wealth would penalise savers relative to spenders, encouraging individuals to save or invest less than they otherwise would. To the extent that the wealthy are internationally mobile, it could also see individuals moving overseas in response to the tax.22 Again, that concern may be particularly salient for the super-wealthy, who may be more able (or more willing) to change their country of residence in response to tax changes.The long-term consequences for economic prosperity would need to be carefully weighed against any distributional benefits that a policymaker considers to be desirable. It is possible that a government might actually want to discourage people from saving to accumulate wealth. But if its real concern is with inequality of lifetime income rather than wealth per se, it would be better to tax sources of wealth once when they are received (and/or when they are spent) rather than taxing the same wealth every year. The UK has taxes on profits, capital income and gains, inheritances and pension saving which are deeply flawed. By reforming the tax base (the definition of what is taxed) as well as tax rates, these taxes could be made to achieve more redistribution and more equal treatment of similarly well-off people while reducing distortions to saving and investment decisions.23 Fixing existing capital taxes might be both easier and better targeted than introducing an annual wealth tax.

Improving the taxation of future income flows – and addressing the ways income is generated in the first place – would take longer to affect the wealth distribution than taxing wealth directly. If the concern is about inequalities in future wealth accumulation – or how it is accumulated – then that is appropriate. But it would not directly address inequalities in the existing stock of wealth. If that is the concern, the most direct tool to redistribute legacy stocks of wealth would be a one-off tax on existing wealth. Defining and valuing taxable wealth would still raise questions of fairness and complexity. But a tax based on an unexpected and credibly one-off assessment of existing wealth could in principle be an economically efficient way to raise revenue, since tax liabilities based only on past wealth could not be reduced by changing future behaviour. The potential efficiency of such a tax could be undermined, however, if announcing a one-off tax created expectations of, or uncertainty about, other future taxes (whether a repeat of the ‘one-off’ wealth tax or something else entirely) and this led people to change their behaviour – though other tax rises can also affect expectations and uncertainty. Whether such a tax is considered fair (e.g. whether it is fair that a one-off tax would fall mainly on the particular generation that happened to be at the peak of their wealth – typically around retirement age – at the time it was assessed, and on those within that generation who had saved rather than spent their money) is something on which reasonable people will differ. Again, it will depend on precisely what the policymaker’s concerns and objectives are: for example, whether they are related to perceived problems with how existing wealth was accrued (or taxed at that time) – though such concerns might not apply equally to all wealth accruals, and current wealth holders will not perfectly correspond to those affected by past events.

Land and property taxation

One alternative to a broad-based wealth tax is a tax on one particular form of wealth: land. This has considerable practical advantages: it is readily identifiable, it cannot move abroad in response to taxation and, because it is essentially in fixed supply, taxing it does not create the disincentive effects associated with other taxes. The UK currently has several property taxes – principally council tax, business rates and stamp duty land tax (and its equivalents in Scotland and Wales), which are badly designed and in dire need of reform even if no additional revenue is to be raised (as explained in Mirrlees et al. (2011)). But raising more revenue from reformed land and property taxation would be an efficient way to tax a wealthy group.24

Whether raising more revenue from property taxation would be fair as well as efficient is more debatable. The fixed supply of land means that its value is reduced pound-for-pound by the value of expected future tax liabilities attached to it (the tax is ‘capitalised’), so increases in property taxation result in a reduction in wealth for the property owners at the time the tax rise is announced. Future purchasers of property would be no worse off as an increase in the tax liability attached to a given property asset would be offset by a reduction in price. As with a one-off wealth tax, therefore, policymakers would need to think carefully about whether it is considered fair to impose a tax on those who were wealthy at that particular point in time. And in this case, they would also need to think carefully about whether it is fair to tax those who happen to hold their wealth in the form of property while someone who is equally wealthy but holds their wealth in non-property assets is left untaxed.

Summing up

Inequalities in wealth can occur for three broad reasons: because people are at different points in the life cycle, because of differences in lifetime income (including inheritances) and because of differences in saving decisions. Many of the concerns people have about wealth inequality – the fact, for instance, that the wealthy are able to enjoy lavish lifestyles while others struggle to get by – are really questions of lifetime income or consumption inequality. If these are policymakers’ concerns, then policy on future wealth accumulation should focus on addressing how incomes are generated and taxed, rather than taxing saving. Policymakers have a wide range of tools available to do that.

It is only if policymakers consider inequality in wealth to be a problem over and above inequality in lifetime income that it provides a potential justification for pursuing more equal accumulation of wealth itself. An annual tax on wealth would directly redistribute wealth, but it would have serious drawbacks on a practical level (in terms of defining and valuing taxable wealth) as well as distorting savings (and potentially location) decisions. Those drawbacks may be especially severe at the very top of the distribution, where wealth is often concentrated in hard-to-value business assets and where international mobility may be greater.

Policy affecting income received and wealth accumulated in future would not address inequalities in existing wealth holdings, which are a result of what happened in the past – sometimes generations ago. Taxes with one-off effects on existing wealth holders – such as a one-off wealth tax or an increase in property taxes – address existing wealth holding directly, and potentially in an economically efficient way, though how far that would be achieved in practice is less clear. How far policymakers ought to try to ‘correct’ for the past by targeting current wealth holders is debatable, and doing so can raise other issues of fairness – for example, between people who have already spent their money and those who still have it, or between people who have their wealth in property and those holding other assets.

Ultimately, what wealth inequalities highlight above all is the paramount importance of precision to successful policymaking. Only by carefully isolating exactly which inequalities are of concern can the right policies be chosen.

4. Conclusion

Many people have significant concerns about inequality or, to be more accurate, about particular inequalities. Where there are concerns, these can have many different roots, and people can often end up talking past each other. Some will predominantly be concerned with improving the circumstances of those at the ‘lower end’ of any particular distribution. Others focus on reining in the ‘top end’ to reduce concentrations of income, wealth or power. Still others are relatively unconcerned about how unequal outcomes are, but do worry about the process of reaching those outcomes: is opportunity widespread, even if success is more concentrated? And another group focuses on the links between different dimensions of inequality: an unequal distribution of educational attainment may not raise eyebrows on its own, but becomes more concerning if it reflects and reinforces pre-existing inequalities by deprivation, geography or ethnicity.

The IFS Deaton Review has provided evidence on these different aspects of inequalities, and more. There is a particular focus on drawing out the relationships between different dimensions of inequality. In some cases, inequalities are in tension with each other. In other areas, inequalities are positively interrelated and work to drive each other. For example, the wealth inequalities discussed in this report link back to the labour market opportunities that people have available. In turn, job opportunities depend partly on where people live in the country: some regions, such as London, have a much higher concentration of high-paying jobs. That is partly driven by the supply of skilled workers, who make decisions about how to develop their skills based on their ability and on the return to their investments. But those decisions are shaped by children’s earliest experiences at home – which are crucially influenced by the wealth of the family they are born into. These outcomes, and inequalities, are connected in a form of cycle.

A policymaker who is concerned about inequalities should take these cycles seriously. Taken together, different dimensions of inequality can become mutually reinforcing: a company choosing where to open its new office might steer clear of an area with few skilled workers already available. Meanwhile, a young person in a disadvantaged region may choose not to invest in additional skills because they do not see a path to a good job that rewards those skills. The resulting lack of a good job has knock-on consequences for their health, life satisfaction and other outcomes.

In some cases, inequalities are a by-product of a market that is not functioning efficiently. If concentrations of wealth at the top end result from market power, that is not just an inequality issue – that means that the economy as a whole does not benefit from an efficient allocation of resources and talent.

In our toolkit, we set out three questions to help policymakers navigate these issues. A policymaker who is concerned about a particular inequality should start by thinking seriously about its causes. Is the inequality the result of a market failure which can be addressed directly? Or is the root cause instead the different choices that people make given their abilities, resources and preferences? The latter case implies harder choices and trade-offs.

The second question asks policymakers to consider the consequences of the inequality of concern – especially for other dimensions of inequality. Does addressing one inequality open up another? Or do these different dimensions reinforce one another? A key conclusion of the IFS Deaton Review is that these different dimensions of inequality cannot easily be considered in isolation.

Finally, we prompt policymakers to consider the costs and benefits of different courses of action – not only for the inequality they are concerned about, but also for both overall efficiency and impacts on other dimensions of inequality. In doing so, policymakers should consider the system as a whole. Not every policy needs to (or can) reduce every inequality, and often the costs will be judged to be too high. At the same time, sometimes this systemic thinking will reveal that the benefits of a course of action are larger than they initially seem. Where market failures are involved, or where multiple inequalities intersect and reinforce, targeting one dimension of inequality can have positive knock-on consequences.

Almost all dimensions of public policy influence, and are influenced by, inequalities. Our toolkit aims to be broad and flexible enough to support a policymaker to navigate addressing whichever dimension of inequality, in whichever policy area, they have identified as a concern. But that choice – to determine what is a concern and in which context – is, and always will be, a political one. People’s preferences are broad and varied, and opinions can shift over time. There is no one right answer here. But what there can be is productive, reasoned discussion based on solid evidence and clear thinking – and that is exactly what this report, and the enormous body of work that it is based on, aim to provide.

Acknowledgements

This report was funded by the Nuffield Foundation, grant reference WEL/43603. The Nuffield Foundation is an independent charitable trust with a mission to advance social well-being. It funds and undertakes rigorous research, encourages innovation and supports the use of sound evidence to inform social and economic policy, and improve people’s lives. The Nuffield Foundation is the founder and co-funder of the Nuffield Council on Bioethics, the Ada Lovelace Institute and the Nuffield Family Justice Observatory. The views expressed are those of the authors and not necessarily of the Nuffield Foundation. We also thank the Economic and Social Research Council for support via the ESRC Centre for the Microeconomic Analysis of Public Policy (grant number ES/Z504634/1). We are grateful to all the researchers whose work contributed to the IFS Deaton Review, in particular the Review panellists. James Banks, Alex Beer, Richard Blundell, Catherine Dennison, Mark Franks, Susan Harkness, Josh Hillman, Paul Johnson, Gavin Kelly, Ruth Maisey, John Pullinger, Anvar Sarygulov, Luke Sibieta, Darcey Snape and Martin Weale provided invaluable comments. Any errors or omissions are the responsibility of the authors alone.

References

Adam, S., Advani, A., Miller, H. and Summers, A., 2024. Capital gains tax reform. In C. Emmerson, P. Johnson and K. Ogden (eds), The IFS Green Budget: October 2024. https://ifs.org.uk/sites/default/files/2024-10/Captial-gains-tax-reform.pdf.

Adam, S., Delestre, I., Emmerson, C. and Sturrock, D., 2023. A blueprint for a better tax treatment of pensions. IFS Report. https://ifs.org.uk/publications/blueprint-better-tax-treatment-pensions.

Adam, S. and Miller, H., 2021. The economic arguments for and against a wealth tax. Fiscal Studies, 42, 457–83. https://doi.org/10.1111/1475-5890.12288.

Adams, A., Balgova, M., Qian, M. and Waters, T., 2023. Firm concentration and job design: the case of schedule flexible work arrangements. IFS Working Paper. https://ifs.org.uk/publications/firm-concentration-job-design-case-schedule-flexible-work-arrangements.

Advani, A., Bangham, G. and Leslie, J., 2021. The UK’s wealth distribution and characteristics of high-wealth households. Fiscal Studies, 42, 397–430. https://doi.org/10.1111/1475-5890.12286.

Advani, A. and Sturrock, D., 2023. Reforming inheritance tax. In C. Emmerson, P. Johnson and B. Zaranko (eds), The IFS Green Budget: October 2023. https://ifs.org.uk/publications/reforming-inheritance-tax.

Advani, A., Summers, A. and Tarrant, H., 2025. Measuring top wealth shares in the UK. European Economic Review, 178, 105076. https://doi.org/10.1016/j.euroecorev.2025.105076.

Aghion, P., Bergeaud, A., Blundell, R. and Griffith, R., 2024. Social skills and the individual wage growth of less educated workers. IFS Working Paper. https://ifs.org.uk/publications/social-skills-and-individual-wage-growth-less-educated-workers-0.

Arold, B. W., Ash, E., MacLeod, W. B. and Naidu, S., 2025. Worker rights in collective bargaining. NBER Working Paper. https://www.nber.org/papers/w33605.

Banks, J. and Berkman, L., 2024. Health. Oxford Open Economics, 3 (Supplement 1), i496–8. https://doi.org/10.1093/ooec/odad093.

Barr, A., Eggleston, J. and Smith, A. A., 2022. Investing in infants: the lasting effects of cash transfers to new families. Quarterly Journal of Economics, 137, 2539–83. https://doi.org/10.1093/qje/qjac023.

Belfield, C., Britton, J., Buscha, F., Dearden, L., Dickson, M., van der Erve, L., Sibieta, L., Vignoles, A., Walker, I. and Zhu, Y., 2018. The impact of undergraduate degrees on early-career earnings. Department for Education Research Report. https://www.gov.uk/government/publications/undergraduate-degrees-labour-market-returns.

Bell, B., Costa, R. and Machin, S., 2022. Why does education reduce crime? Journal of Political Economy, 130, 732–65. https://doi.org/10.1086/717895.

Benach, J., Vives, A., Amable, M., Vanroelen, C., Tarafa, G. and Muntaner, C., 2014. Precarious employment: understanding an emerging social determinant of health. Annual Review of Public Health, 35, 229–53. https://www.annualreviews.org/doi/10.1146/annurev-publhealth-032013-182500.

Benmelech, E., Bergman, N. K. and Kim, H., 2022. Strong employers and weak employees: how does employer concentration affect wages? Journal of Human Resources, 57 (S), S200–50. https://jhr.uwpress.org/content/57/S/S200.

Benson, R., Duffy, B., Hesketh, R. and Hewlett, K., 2024. Attitudes to inequalities. Oxford Open Economics, 3 (Supplement 1), i39–63. https://doi.org/10.1093/ooec/odad069.

Blanden, J., Greaves, E., Gregg, P., Macmillan, L. and Sibieta, L., 2015. Understanding the improved performance of disadvantaged pupils in London. CASE Working Paper. https://sticerd.lse.ac.uk/CASE/_NEW/PUBLICATIONS/abstract/?index=4750.

Blundell, R., Costa Dias, M., Goll, D. and Meghir, C., 2021. Wages, experience, and training of women over the life cycle. Journal of Labor Economics, 39, S275–315. https://www.journals.uchicago.edu/doi/abs/10.1086/711400.

Blundell, R., Costa Dias, M., Meghir, C. and Shaw, J., Female labor supply, human capital, and welfare reform. Econometrica, 84, 1705–53. https://onlinelibrary.wiley.com/doi/abs/10.3982/ECTA11576.

Blundell, R., Cribb, J., Wernham, T. and Xu, X., 2023. Inequality in the UK: 1968–2021. Country Studies: Inequalities in Europe and North America – a parallel study to the IFS Deaton Review.  https://ifs.org.uk/publications/inequality-uk-1968-2021.

Blundell, R., Dearden, L. and Meghir, C., 1996. The determinants and effects of work-related training in Britain. IFS Report. https://ifs.org.uk/publications/determinants-and-effects-work-related-training-britain.

Bourquin, P., Brewer, M. and Wernham, T., 2024. Trends in income and wealth inequalities. Oxford Open Economics, 3, Issue Supplement_1, i103–46,  https://doi.org/10.1093/ooec/odad100.

Bourquin, P., Joyce, R. and Sturrock, D., 2021. Inheritances and inequality over the life cycle: what will they mean for younger generations? IFS Report. https://ifs.org.uk/publications/inheritances-and-inequality-over-life-cycle-what-will-they-mean-younger-generations.

Brandily, P., Distefano, M., Donnat, H., Overman, H. G. and Shah, K., 2022. Bridging the gap: what would it take to narrow the UK’s productivity disparities? Resolution Foundation, Economy 2030 Inquiry Report,  https://economy2030.resolutionfoundation.org/reports/bridging-the-gap/.

Breen, R., 2024. Persistent inequality: contemporary inequality in a historical context. Oxford Open Economics, 3, Issue Supplement_1, i821–7, https://doi.org/10.1093/ooec/odad015.

Brewer, M. and Hoynes, H., 2020. In-work credits in the UK and the US. Fiscal Studies, 40, 519–60. https://onlinelibrary.wiley.com/doi/10.1111/1475-5890.12207.

Briole, S., Gurgand, M., Maurin, E., McNally, S., Ruiz-Valenzuela, J. and Santín, D., 2025. The making of civic virtues: a school-based experiment in three countries. American Economic Journal: Policy, 17, 496–521. https://www.aeaweb.org/articles?id=10.1257/pol.20230612.

Britton, J., Dearden, L. and Waltmann, B., 2021. The returns to undergraduate degrees by socio-economic group and ethnicity. IFS Report. https://ifs.org.uk/publications/returns-undergraduate-degrees-socio-economic-group-and-ethnicity.

Britton, J., van der Erve, L., Waltmann, B. and Xu, X., 2021. London calling? Higher education, geographical mobility and early-career earnings. Department for Education Research Report. https://doi.org/10.1920/re.ifs.2021.0198.

Browning, M. and Heinesen, E., 2012. Effect of job loss due to plant closure on mortality and hospitalization. Journal of Health Economics, 31, 599–616. https://doi.org/10.1016/j.jhealeco.2012.03.001.

Burgess, S., Cantillon, E., Cavallo, M. and Greaves, E., 2025. Modifying school choice for more equitable access in England. University of Bristol. https://www.bristol.ac.uk/media-library/sites/economics/documents/modifying-school-choice-for-more-equitable-access-in-england-2025.pdf.

Burgess, S., Cantillon, E., Cavallo, M., Greaves, E. and Zhang, M., 2023. School admissions in England: the rules schools choose on which pupils to admit. ULB Institutional Repository 2013/356676, https://ideas.repec.org/p/ulb/ulbeco/2013-356676.html.

Burgess, S., Greaves, E. and Vignoles, A., 2019. School choice in England: evidence from national administrative data. Oxford Review of Education, 45, 690–710. https://doi.org/10.1080/03054985.2019.1604332.

Burgess, S., Greaves, E., Vignoles, A. and Wilson, D., 2015. What parents want: school preferences and school choice. Economic Journal, 125, 1262–89. https://doi.org/10.1111/ecoj.12153.

Bursztyn, L., González, A. L. and Yanagizawa-Drott, D., 2020. Misperceived social norms: women working outside the home in Saudi Arabia. American Economic Review, 110, 2997–3029. https://doi.org/10.1257/aer.20180975.

Burt, S. A., 2009. Rethinking environmental contributions to child and adolescent psychopathology: a meta-analysis of shared environmental influences. Psychological Bulletin, 135, 608–37. https://doi.org/10.1037/a0015702.

Buscha, F. and Dickson, M., 2015. The wage returns to education over the life-cycle: heterogeneity and the role of experience. IZA Discussion Paper 9596, https://www.iza.org/publications/dp/9596/the-wage-returns-to-education-over-the-life-cycle-heterogeneity-and-the-role-of-experience.

Card, D., Kluve, J. and Weber, A., 2018. What works? A meta analysis of recent active labor market program evaluations. Journal of the European Economic Association, 16, 894–931. https://academic.oup.com/jeea/article/16/3/894/4430618.

Carneiro, P., Cattan, S., Conti, G., Crawford, C., Farquharson, C. and Ridpath, N., 2025a. The short- and medium-term effects of Sure Start on children’s outcomes. IFS Report, https://ifs.org.uk/publications/short-and-medium-term-effects-sure-start-childrens-outcomes.

Carneiro, P., Foliano, F., Krutikova, S., Loh, J. and Macmillan, L., 2025b. Not just a North–South divide: the geography of opportunity in England. Fiscal Studies, 46, 423–50, https://doi.org/10.1111/1475-5890.70012.

Case, A. and Kraftman, L., 2024. Health inequalities. Oxford Open Economics, 3, Issue Supplement_1, i499–528, https://doi.org/10.1093/ooec/odad060.

Cattan, S., Conti, G., Farquharson, C., Ginja, R. and Pecher, M., 2025. The health effects of universal early childhood interventions: evidence from Sure Start. IFS Working Paper. https://ifs.org.uk/publications/health-effects-universal-early-childhood-interventions-evidence-sure-start-0.

Cattan, S., Fitzsimons, E., Goodman, A., Phimister, A., Ploubidis, G. B. and Wertz, J., 2024. Early childhood inequalities. Oxford Open Economics, 3 (Supplement 1), i711–40. https://doi.org/10.1093/ooec/odad072.

Cattan, S., Waters, T. and Wernham, T., 2025. What is the effect of the two-child limit on children’s school readiness? IFS Report, https://ifs.org.uk/publications/what-effect-two-child-limit-childrens-school-readiness.

Chandola, T., Brunner, E. and Marmot, M., 2006. Chronic stress at work and the metabolic syndrome: prospective study. BMJ, 332, 521–5. https://www.bmj.com/content/332/7540/521.long.

Chen, H., Gompers, P., Kovner, A. and Lerner, J., 2009. Buy local? The geography of venture capital. Journal of Urban Economics, 67, 90–102. https://doi.org/10.1016/j.jue.2009.09.013.

Clark, D. and Royer, H., 2013. The effect of education on adult mortality and health: evidence from Britain. American Economic Review, 103, 2087–120. https://www.aeaweb.org/articles?id=10.1257/aer.103.6.2087.

Cominetti, N., McCurdy, C., Thwaites, G. and Vieira-Marques, R., 2023. Low pay Britain 2023: improving low-paid work through higher minimum standards. Resolution Foundation, Economy 2030 Inquiry Report,  https://economy2030.resolutionfoundation.org/reports/low-pay-britain-2023.

Cornelissen, T., Dustmann, C., Raute, A. and Schönberg, U., 2018. Who benefits from universal child care? Estimating marginal returns to early child care attendance. Journal of Political Economy, 126, 2356–409. https://doi.org/10.1086/699979.

Crawford, C., Edwards, A., Farquharson, C., Greaves, E., Trevelyan, G., Wallace, E. and White, C., 2019. Magic Breakfast: evaluation report and executive summary.  https://d2tic4wvo1iusb.cloudfront.net/production/documents/projects/Magic_Breakfast_report.pdf?v=1775899482.

Cribb, J. and Emmerson, C., 2020. What happens to workplace pension saving when employers are obliged to enrol employees automatically? International Tax and Public Finance, 27, 664–93. https://doi.org/10.1007/s10797-019-09565-6.

Daams, M. N., Mayer, C. and McCann, P., 2025. Regions, cities and finance: the role of capital shocks and banking reforms in shaping the UK geography of prosperity. Fiscal Studies, 46, 475–90. https://doi.org/10.1111/1475-5890.70002.

Daly, S., Hughson, H. and Loutzenhiser, G., 2021. Valuation for the purposes of a wealth tax. Fiscal Studies, 42, 615–50. https://doi.org/10.1111/1475-5890.12287.

Datta, N., Giupponi, G. and Machin, S., 2019. Zero-hours contracts and labour market policy. Economic Policy, 34, 369–427. https://academic.oup.com/economicpolicy/article-abstract/34/99/369/5524663.

Davies, N. M., Dickson, M., Davey Smith, G., van den Berg, G. J. and Windmeijer, F., 2018. The causal effects of education on health outcomes in the UK Biobank. Nature Human Behaviour, 2, 117–25. https://www.nature.com/articles/s41562-017-0279-y.

Delestre, I., Jacobs-Strom, B., Miller, H. and Smith, K., 2025. New trends in self-employment and top incomes. IFS Report. https://ifs.org.uk/publications/new-trends-self-employment-and-top-incomes.

Delestre, I., Kopczuk, W., Miller, H. and Smith, K., 2024. Top income inequality and tax policy. Oxford Open Economics, 3, Issue Supplement_1, i1086–112, https://doi.org/10.1093/ooec/odad068.

De Loecker, J., Obermeier, T. and Van Reenen, J., 2024. Firms and inequality. Oxford Open Economics, 3, Issue Supplement_1, i962–82,  https://doi.org/10.1093/ooec/odad097.

Diamond, R., 2016. The determinants and welfare implications of US workers’ diverging location choices by skill: 1980–2000. American Economic Review, 106, 479–524. https://doi.org/10.1257/aer.20131706.

Diniz, J. and Murphy, L., 2025. False starts: what the UK’s growing NEETs problem really looks like, and how to fix it. Resolution Foundation, https://www.resolutionfoundation.org/app/uploads/2025/10/False-starts.pdf.

Doepke, M. and Zilibotti, F., 2019. Love, Money & Parenting: How Economics Explains the Way We Raise Our Kids. Princeton University Press.

Dube, A. and Lindner, A., 2024. Minimum wages in the 21st century. NBER Working Paper. https://www.nber.org/papers/w32878.

Duranton, G. and Puga, D., 2004. Micro-foundations of urban agglomeration economies. In J. V. Henderson and J. F. Thisse (eds), Handbook of Regional and Urban Economics, vol. 4. Elsevier.

Emmerson, C. and Wakefield, M., 2001. The Saving Gateway and the Child Trust Fund: is asset-based welfare ‘well fair’? IFS Report. https://ifs.org.uk/publications/saving-gateway-and-child-trust-fund-asset-based-welfare-well-fair.

Espinoza, H., Speckesser, S., Tahir, I., Britton, J., McNally, S. and Vignoles, A., 2020. Post-18 education: who is taking different routes and how much do they earn? Centre for Vocational Education Research. https://cver.lse.ac.uk/_new/publications/abstract/?index=7270.

Eurofound, 2020. Budapest, Paris and Amsterdam report longest commuting times among EU capitals. Press release, 13 February,  https://www.mynewsdesk.com/eurofound/news/budapest-paris-and-amsterdam-report-longest-commuting-times-among-eu-capitals-394279.

Eurostat, 2026. Average rent per month in cities by type of dwelling [prc_colc_rents]. https://ec.europa.eu/eurostat/databrowser/view/prc_colc_rents/default/table?lang=en.

Fagereng, A., Guiso, L., Malacrino, D. and Pistaferri, L., 2020. Heterogeneity and persistence in returns to wealth. Econometrica, 88, 115–70. https://doi.org/10.3982/ECTA14835.

Faggio, G., Silva, O. and Strange, W. C., 2017. Heterogeneous agglomeration. Review of Economics and Statistics, 99, 80–94. https://doi.org/10.1162/REST_a_00604.

Farquharson, C., McNally, S. and Tahir, I., 2024. Education inequalities. Oxford Open Economics, 3, Issue Supplement_1, i760–820, https://doi.org/10.1093/ooec/odad029.

Farquharson, C., Ogden, K., Ornadel, N., Ridpath, N., Sibieta, L., Snape, D. and Tahir, I., 2026. Annual report on education spending in England: 2025–26. IFS Report. https://ifs.org.uk/publications/annual-report-education-spending-england-2025-26.

Frick, S., Taylor, I., Prenzel, P., Penney, K., Collier, P., Goodstadt, V., Mayer, C. and McCann, P., 2023. Lessons from successful ‘turnaround’ cities for the UK. Resolution Foundation, Economy 2030 Inquiry Report, https://economy2030.resolutionfoundation.org/reports/lessons-from-successful-turnaround-cities-for-the-uk/.

Gibbons, S., Machin, S. and Silva, O., 2013. Valuing school quality using boundary discontinuities. Journal of Urban Economics, 75, 15–28. https://doi.org/10.1016/j.jue.2012.11.001.

Giupponi, G., Joyce, R., Lindner, A., Waters, T., Wernham, T. and Xu, X., 2024. The employment and distributional impacts of nationwide minimum wage changes. Journal of Labor Economics, 42, S293–333. https://doi.org/10.1086/728471.

Giupponi, G., Ray-Chaudhuri, S. and Xu, X., 2024. How has the NLW affected pay differentials within firms? IFS Comment. https://ifs.org.uk/articles/how-has-nlw-affected-pay-differentials-within-firms.

Giupponi, G. and Xu, X., 2020. What does the rise of self-employment tell us about the UK labour market? IFS Report, https://ifs.org.uk/publications/what-does-rise-self-employment-tell-us-about-uk-labour-market.

Goldin, J. and Reck, D., 2022. Optimal defaults with normative ambiguity. Review of Economics and Statistics, 104, 17–33. https://direct.mit.edu/rest/article-abstract/104/1/17/97651/Optimal-Defaults-with-Normative-Ambiguity.

Gravoueille, M., 2025. Wage and employment effects of wage subsidies. https://www.maximegravoueille.com/research/IncidenceWageSubsidies.pdf.

Gruber, J., 2008. Mandated employer provision of employee benefits. In S. N. Durlauf and L. E. Blume (eds), The New Palgrave Dictionary of Economics. https://doi.org/10.1007/978-1-349-58802-2_1014.

Hayward, H., Hunt, E. and Lord, A., 2014. The economic value of key intermediate qualifications: estimating the returns and lifetime productivity gains to GCSEs, A levels and apprenticeships. Department for Education, Research Report, https://www.gov.uk/government/publications/gcses-a-levels-and-apprenticeships-their-economic-value.

Hendren, N. and Sprung-Keyser, B., 2020. A unified welfare analysis of government policies. Quarterly Journal of Economics, 135, 1209–318, https://doi.org/10.1093/qje/qjaa006.

Higher Education Statistics Agency, 2022. HE students by sex, level of study, mode of study and academic year. https://www.hesa.ac.uk/data-and-analysis/students/chart-3.

Institute for the Future of Work, 2025. Final Report of the Pissarides Review into the Future of Work and Wellbeing. https://www.ifow.org/publications/the-final-report-of-the-pissarides-review.

Jakobsen, K., Kleven, H., Kolsrud, J., Landais, C. and Muñoz, M., 2025. Taxing top wealth: migration responses and their aggregate economic implications. https://econ.lse.ac.uk/staff/clandais/cgi-bin/Articles/WealthMig_R.pdf.

Johnson, P., Banks, J., Besley, T., Blundell, R., Deaton, A., Joyce, R. and Satz, D., 2026. Challenging Inequalities: How We Got Stuck and Where We Go Next. Princeton University Press.

Katz, L. F., Roth, J., Hendra, R. and Schaberg, K., 2022. Why do sectoral employment programs work? Lessons from WorkAdvance. Journal of Labor Economics, 40, S249–91, https://www.journals.uchicago.edu/doi/abs/10.1086/717932.

Kline, P. and Moretti, E., 2014. Local economic development, agglomeration economies, and the big push: 100 years of evidence from the Tennessee Valley Authority. Quarterly Journal of Economics, 129, 275–331, https://doi.org/10.1093/qje/qjt034.

Kopczuk, W. and Saez, E., 2004. Top wealth shares in the United States, 1916–2000: evidence from estate tax returns. National Tax Journal, 57, 445–87, https://doi.org/10.17310/ntj.2004.2S.05. 

Kovalenko, A., 2023. Natural resource booms, human capital, and earnings: evidence from linked education and employment records. American Economic Journal: Applied Economics, 15, 184–217, https://doi.org/10.1257/app.20200762.

Krause, P. K., Rhodes, E., Miller, S., Bartik, A. W., Broockman, D. E. and Vivalt, E., 2026. The impact of unconditional cash transfers on parenting and children. NBER Working Paper, https://www.nber.org/papers/w34040.

Kuhn, M., Schularick, M. and Steins, U. I., 2020. Income and wealth inequality in America, 1949–2016. Journal of Political Economy, 128, 3469–519, https://doi.org/10.1086/708815.

Latimer, E., Sibieta, L. and Snape, D., 2025. Support for children with disabilities and special educational needs. IFS Report, https://ifs.org.uk/publications/support-children-disabilities-and-special-educational-needs.

Levell, P. and Sturrock, D., 2026. How do house prices affect social mobility? IFS Working Paper, https://ifs.org.uk/publications/how-do-house-prices-affect-social-mobility.

Lipsey, R. G. and Lancaster, K., 1956. The general theory of second best. Review of Economic Studies, 24, 11–32, https://doi.org/10.2307/2296233.

Lleras-Muney, A., 2005. The relationship between education and adult mortality in the United States. Review of Economic Studies, 72, 189–221, https://doi.org/10.1111/0034-6527.00329. 

Machin, S., McNally, S. and Ruiz-Valenzuela, J., 2020. Entry through the narrow door: the costs of just failing high stakes exams. Journal of Public Economics, 190, 104224, https://doi.org/10.1016/j.jpubeco.2020.104224.

Maestas, N., Mullen, K. J., Powell, D., von Wachter, T. and Wenger, J. B., 2023. The value of working conditions in the United States and the implications for the structure of wages. American Economic Review, 113, 2007–47, https://www.aeaweb.org/articles?id=10.1257/aer.20190846.

Manning, A., 2001. Monopsony and the efficiency of labour market interventions. CEP Discussion Paper, https://cep.lse.ac.uk/_new/publications/abstract.asp?index=514.

Mas-Colell, A., Whinston, M. D. and Green, J. R., 1995. Microeconomic Theory. Oxford University Press.

McCann, P., 2020. Perceptions of regional inequality and the geography of discontent: insights from the UK. Regional Studies, 54, 256–67, https://doi.org/10.1080/00343404.2019.1619928.

McLean, D. and Worth, J., 2023. Impact evaluation of the Teach First training programme. National Foundation for Education Research, https://www.nfer.ac.uk/publications/impact-evaluation-of-the-teach-first-training-programme/.

Meaburn, E. L., 2025. Navigating genomics and education: insights, opportunities and challenges. Nuffield Council on Bioethics and Nuffield Foundation, https://www.nuffieldbioethics.org/publication/navigating-genomics-and-education-insights-opportunities-and-challenges/.

Mirrlees, J., Adam, S., Besley, T., Blundell, R., Bond, S., Chote, R., Gammie, M., Johnson, P., Myles, G. and Poterba, J., 2011. Tax by Design: The Mirrlees Review, Oxford University Press for the Institute for Fiscal Studies, https://ifs.org.uk/books/tax-design.

Moretti, E., 2024. Place-based policies and geographical inequalities. Oxford Open Economics, 3, Issue Supplement_1, i625–33, https://doi.org/10.1093/ooec/odad049.

Moretti, E. and Wilson, D. J., 2023. Taxing billionaires: estate taxes and the geographical location of the ultra-wealthy. American Economic Journal: Economic Policy, 15, 424–66, https://doi.org/10.1257/pol.20200685.

Neumark, D. and Simpson, H., 2015. Place-based policies. In G. Duranton, J. V. Henderson and W. C. Strange (eds), Handbook of Regional and Urban Economics, volume 5, Elsevier.

Noble, K., Magnuson, K., Duncan, G., Gennetian, L. A., Yoshikawa, H., Fox, N. A., Halpern-Meekin, S., Troller-Renfree, S., Han, S., Egan-Dailey, S., Nelson, T. D., Mize Nelson, J., Black, S., Georgieff, M. and Karhson, D., 2025. The effect of a monthly unconditional cash transfer on children’s development at four years of age: a randomized controlled trial in the U.S. NBER Working Paper, https://www.nber.org/papers/w33844. 

Office for National Statistics, 2026. Healthy life expectancy, UK: between 2011 to 2013 and 22 to 2024. https://www.ons.gov.uk/peoplepopulationandcommunity/healthandsocialcare/healthandlifeexpectancies/bulletins/healthstatelifeexpectanciesuk/between2011to2013and2022to2024.

Oreopoulos, P. and Salvanes, K. G., 2011. Priceless: the nonpecuniary benefits of schooling. Journal of Economic Perspectives, 25, 159–84, https://www.aeaweb.org/articles?id=10.1257/jep.25.1.159.

Overman, H. G. and Xu, X., 2024. Spatial disparities across labour markets. Oxford Open Economics, 3, Issue Supplement_1, i585–610, https://doi.org/10.1093/ooec/odae005.

Oxley, F. A. R., Wilding, K. and von Stumm, S., 2024. DNA and IQ: big deal or much ado about nothing? A meta-analysis. Intelligence, 107, 101871, https://doi.org/10.1016/j.intell.2024.101871. 

Page, M. E., 2024. New advances on an old question: does money matter for children’s outcomes? Journal of Economic Literature, 62, 891–947, https://www.aeaweb.org/articles?id=10.1257/jel.20231553.

Pierce, J. R. and Schott, P. K., 2020. Trade liberalization and mortality: evidence from US counties. American Economic Review: Insights, 2, 47–64, https://doi.org/10.1257/aeri.20180396.

Piketty, T. and Zucman, G., 2014. Capital is back: wealth-income ratios in rich countries, 1700–2010. Quarterly Journal of Economics, 129, 1255–310, https://doi.org/10.1093/qje/qju018.

Reader, M., 2023. The infant health effects of starting universal child benefits in pregnancy: evidence from England and Wales. Journal of Health Economics, 89, 102751, https://doi.org/10.1016/j.jhealeco.2023.102751.

Robling, M., Lugg-Widger, F. V., Cannings-John, R., Angel, L., Channon, S., Fitzsimmons, D., Hood, K., Kenkre, J., Moody, G., Owen-Jones, E., Pockett, R. D., Sanders, J., Segrott, J. and Slater, T., 2022. Nurse-led home-visitation programme for first-time mothers in reducing maltreatment and improving child health and development (BB:2-6): longer-term outcomes from a randomised cohort using data linkage. BMJ Open, 12, e049960, https://bmjopen.bmj.com/content/12/2/e049960.

Rodrik, D. and Stantcheva, S., 2021a. Fixing capitalism’s good jobs problem. Oxford Review of Economic Policy, 37, 824–37, https://academic.oup.com/oxrep/article/37/4/824/6423487.

Rodrik, D. and Stantcheva, S., 2021b. A policy matrix for inclusive prosperity. NBER Working Paper, https://doi.org/10.3386/w28736.

Rothstein, J., 2010. Is the EITC as good as an NIT? Conditional cash transfers and tax incidence. American Economic Journal: Economic Policy, 2, 177–208, https://www.aeaweb.org/articles?id=10.1257/pol.2.1.177.

Saez, E. and Zucman, G., 2016. Wealth inequality in the United States since 1913: evidence from capitalized income tax data. Quarterly Journal of Economics, 131, 519–78, https://doi.org/10.1093/qje/qjw004.

Sibieta, L., 2021. School spending in England: trends over time and future outlook. IFS Report, https://ifs.org.uk/publications/school-spending-england-trends-over-time-and-future-outlook.

Silva, A. and Wessel, M., 2025. How to connect communities to create opportunities. Behavioural Insights Team, https://www.bi.team/publications/social-capital-in-the-united-kingdom/.

Smith, K. and Miller, H., 2025. The efficiency and equity effects of capital tax reform. https://katesmith.me/wp-content/uploads/2025/01/smithmiller2025.pdf.

Smith, M., Zidar, O. and Zwick, E., 2023. Top wealth in America: new estimates under heterogeneous returns. Quarterly Journal of Economics, 138, 515–73, https://doi.org/10.1093/qje/qjac033.

Stansbury, A., Turner, D. and Balls, E., 2023. Tackling the UK’s regional economic inequality: binding constraints and avenues for policy intervention. Contemporary Social Science, 18, 318–56, https://doi.org/10.1080/21582041.2023.2250745.

Tirole, J., 2024. Competition and industrial policy in the 21st century. Oxford Open Economics, 3, Issue Supplement_1, i983–1001, https://doi.org/10.1093/ooec/odad080.

Varian, H., 2014. Intermediate Economics: A Modern Approach. Eighth edition, W. W. Norton & Company, Inc.

Vermeulen, P., 2018. How fat is the top tail of the wealth distribution? Review of Income and Wealth, 64, 357–87, https://doi.org/10.1111/roiw.12279.

Vignoles, A., 2016. What is the economic value of literacy and numeracy? IZA World of Labor, https://www.econstor.eu/bitstream/10419/148454/1/iza-wol-229.pdf.

Worth, J. and McLean, D., 2025. The impact of early career retention payments on teacher retention. National Foundation for Education Research. https://www.nfer.ac.uk/publications/the-impact-of-early-career-retention-payments-on-teacher-retention/.

Xu, X., 2023. The changing geography of jobs. IFS Report, https://ifs.org.uk/publications/changing-geography-jobs.