Taxes on wealth: time for another look?

Around the world, proposals for new taxes on the stock of wealth (a ‘wealth tax’) were gaining traction even before the fiscal tumult that has followed the COVID-19 pandemic. Piketty's ‘utopian ideal’ of a global wealth tax is perhaps the most prominent example.11 Piketty, 2014. In the US, a new wealth tax proposed by Saez and Zucman (2019) was taken up by prominent candidates for the 2019 Democratic primaries. The OECD recently conducted a wide-ranging review of the wealth taxes already in operation within its member countries,22 OECD, 2018. and new empirical research has also examined the impacts of existing wealth taxes in a developing country context.33 Londoño-Vélez and Ávila-Mahecha, 2021. The COVID-19 crisis has accelerated this interest in wealth taxes. First off the mark was Argentina, which in December 2020 passed a new one-off levy on its wealthiest citizens (with assets over $2.5 million) to pay for medical supplies and relief measures.44 https://www.bbc.co.uk/news/world-latin-america-55199058. In April 2021, in response to concerns about a $5 trillion surge in the wealth of the world's richest individuals during the pandemic, the UN Secretary-General urged governments to consider introducing a new ‘solidarity or wealth tax’.55 https://www.reuters.com/article/us-health-coronavirus-un-idUSKBN2BZ281. The International Monetary Fund has also recommended that countries consider taxes on high wealth, emphasising the ‘symbolic impact of this type of contribution’ in the context of recovery from the COVID-19 crisis.66 https://www.ft.com/content/5dad2390-8a32-4908-8c96-6d23cd037c38. These developments follow several decades in which taxes on the stock of wealth were almost entirely off the global political agenda. In fact, starting in the 1990s, many of the countries that already had an annual wealth tax, decided to abolish them.77 OECD, 2018. These examples have fed a narrative that whilst wealth taxes may be popular with the public in principle, they cannot be made to work in practice. Only three countries within the OECD still have an annual wealth tax: Norway, Spain and Switzerland. The design of the tax in each of these countries differs quite significantly, although none operates the high-threshold broad-based approach that has been advocated by Piketty, Saez, Zucman and others of late. In the UK, it has been nearly half a century since proposals for a wealth tax were last seriously evaluated. In the aftermath of World War II, Nicholas Kaldor recommended imposing an annual wealth tax as a replacement for the income ‘surtax’.88 Kaldor, 1951. Kaldor later served on the Royal Commission on Taxation 1951–55, which rejected the idea of a wealth tax, together with Kaldor's better-known proposals for an expenditure tax.99 HMSO, 1955. The Labour Party promised to introduce an annual wealth tax in their 1974 manifesto and issued a Green Paper once in government.1010 HMSO, 1974. These plans attracted criticism from the newly formed Institute for Fiscal Studies, amongst others.1111 See Glennerster (2012), although Sandford, Willis and Ironside (1975), all affiliates of the IFS, published a more positive review of the case for an annual wealth tax at the time. Following doubts by Treasury officials and staunch opposition from external groups,1212 Glennerster, 2012. Labour's plans for a wealth tax were finally shelved in 1976. ‘wealth gives opportunity, security, social power, influence and independence. For this reason it may be argued that, however well a system of taxation of income or of consumption may be devised, equity requires that wealth itself should be included in the base for progressive taxation.’ Despite making the ‘in-principle’ case for a wealth tax, the Meade Report did not develop a specific proposal and the idea was not taken any further. In the 30 years that followed, through the 1980s up until the global financial crisis of 2007–08, one can find hardly anything at all written about the idea of introducing an annual wealth tax in the UK. The contributors to the Mirrlees Review, published in 2011, briefly considered a regular wealth tax, but only to reject it.1414 Boadway, Chamberlain and Emmerson, 2010. The review proceeded – unlike the Meade Report – on the basis that the only benefit individuals receive from holding wealth comes from the consumption (by the individual or whoever the wealth is passed on to) that the wealth can finance.1515 See Banks and Diamond (2010). For further dicussion, see Adam and Miller (2021) and Summers (2021) in this issue. In rejecting an annual wealth tax, the review also cited practical difficulties and poor international experience: ‘[i]t is costly to administer, might raise little revenue, and could operate unfairly and inefficiently’.1616 Boadway et al., 2010. This potted history discloses two important reasons for refreshing thinking on a wealth tax. First, the in-principle case for a wealth tax remains contingent and contested. The issue that divided the Meade Report and the Mirrlees Review is still live. Second, the question of whether and how a wealth tax could work in practice in the UK has received hardly any attention since the 1970s. In the intervening period, the UK's economic circumstances have changed significantly, and so has the landscape of tax administration. As other countries have introduced – and, in many cases, abandoned – a wealth tax, more evidence has also become available from which to draw comparisons and learn lessons. For both of these reasons, it is wise to draw on the ‘state of the art’ rather than relying on old tropes. There is also a third reason for reinvigorating debates about a wealth tax now, in the aftermath of the COVID-19 pandemic. Countries around the world are facing huge pressures on their public finances. These result not only from the increased public spending and reduced tax base during the acute phase of the pandemic, but also from calls for sustainably higher spending in domains where the cracks have grown largest, such as in health, social care and education. These circumstances provide a renewed urgency to ‘think big’ on tax policy and to treat seriously, albeit with due scrutiny, options that were previously considered off the table. The lack of a robust and up-to-date evidence base on wealth taxes risks leaving policymakers unprepared for this critical debate. This special issue draws together the latest thinking on wealth taxes with the aim of filling this gap. It draws heavily on international experience and evidence, applying these insights to the UK context. The papers build on work undertaken for the Wealth Tax Commission, which delivered its final report in December 2020.1717 Advani, Chamberlain and Summers, 2020. The contributors to this special issue include tax practitioners and academic lawyers as well as economists, reflecting our view that this range of expertise is essential to evaluating the practice, as well as principles, of a wealth tax. In the following section, we briefly touch upon some common themes arising across the papers. We also highlight some important remaining gaps in the evidence base on wealth taxes, particularly on the measurement of wealth and behavioural responses at the very top of the wealth distribution. A logical first step in asking whether a wealth tax would be desirable is to consider the in-principle case for and against. This is the focus of Adam and Miller (2021), who review the arguments from economic theory. They draw an important distinction – reflected in several papers in this special issue, as well as in the Wealth Tax Commission's final report – between a one-off wealth tax and a recurring (typically annual) tax. Broadly, Adam and Miller are more supportive of the former than the latter, although in both cases it is difficult to stick strictly to conclusions based on theory alone, because (as they emphasise) questions of fairness, which are less amenable to economic analysis, also enter into consideration – especially for a one-off tax – as well as practical concerns. Summers (2021) also addresses the principled case for and against a wealth tax, but in the context of the UK's existing taxes on wealth (broadly defined). He outlines two possible directions for reform, designed around either consumption or income as the primary tax base. A tax on the stock of wealth could not achieve the objectives of either set of reforms, so is not a viable alternative. However, Summers suggests that an additional annual tax on extreme wealth might be justified if the aim was to curb top-end inequality within a short timescale, due to its capacity – unlike taxes on the returns to wealth or on inheritances – to erode individuals’ current stock of assets without waiting for them to make a disposal (or die). Often in tax design, principles give way to political expedience. The political dimension of wealth taxes is picked up by several of the papers in this special issue. Rowlingson, Sood and Tu (2021) report survey evidence that a wealth tax would be more popular with the electorate than other revenue-equivalent tax rises. However, we see polling as more informative on the general priorities and concerns of voters – in this case, what they want from the tax system – than as a barometer for specific policies, especially when those policies (like a wealth tax) are unfamiliar. In this respect, it is notable that Rowlingson et al. (2021) found that the public's most-favoured reason for a wealth tax was their concern with ‘the gap between rich and poor’ and – relatedly – their concern that the richest would not actually pay, which emerged as the strongest argument against. Perret (2021) shows that this worry about loopholes – in the form of exemptions and reliefs – chimes with international experience of annual wealth taxes across Europe. She studies why many countries have abandoned their wealth tax in recent years. Their decline appears to be characterised by a vicious cycle: lobbying leads to design flaws, which leads to predictable avoidance, which undermines the efficiency and fairness of the tax and (critically) the revenue that it can raise. Eventually, the tax raises so little, and distorts so much, that it becomes dispensable and not worth the political contention. Indeed, in the UK, we already see a similar pattern with inheritance tax. A key question, then, is whether in practice a new annual wealth tax could survive the pressure for exemptions and reliefs that have dogged other taxes on wealth. O'Donovan (2021) looks at past experience with various forms of one-off wealth tax. He charts the history of these measures – typically in response to major wars and crises – showing that many were implemented successfully and raised significant revenue. However, these taxes were still not immune from the political difficulties identified by Perret (2021). O'Donovan's survey is necessarily skewed towards experience from the more distant past, as very few recent examples exist. Whereas several countries implemented broad-based one-off wealth taxes in the aftermath of World War II, following the 2008 financial crisis, only Iceland (whose economy was amongst the hardest hit) did so. It remains to be seen whether the COVID-19 crisis will lead to any greater uptake. Chamberlain (2021) makes a compelling case that any wealth tax – whether annual or one-off – must adopt a comprehensive tax base encompassing all types of property, including – most controversially, from a political perspective – main residences, private pensions and business wealth. There is obviously a strong economic case for this approach to limit distortions to savings and investment decisions (at least across asset types). But Chamberlain approaches the matter from a legal perspective and makes the case that a comprehensive base is also the only way to limit avoidance. She further emphasises that exemptions for certain types of asset undermine horizontal equity between taxpayers who have the same taxable capacity but hold their wealth in different forms. It must be acknowledged (indeed, signposted with red flashing lights) that no actual wealth tax has ever achieved a comprehensive base in practice. There are two key practical challenges with implementing an effective wealth tax – especially one with a comprehensive base. First, how should assets be valued? Second, how should we deal with taxpayers who face liquidity constraints, or in other words are ‘asset rich, cash poor’? Both of these issues are often touched upon in the existing economic literature on wealth taxes, but detailed engagement has been notably lacking. Two papers in this special issue seek to remedy this gap: Daly, Hughson and Loutzenhiser (2021) and Loutzenhiser and Mann (2021). Both papers combine empirical estimates of the scale of the issue with a legal perspective on the available policy responses. Whilst the estimates necessarily rely on imperfect data, they initiate the kind of analysis that will be essential if plans for a wealth tax are ever to transition from broad aspiration to workable policy. Daly et al. (2021) draw attention to the fact that (in the UK) valuation on an ‘open market’ basis is already required for several existing taxes, even in the absence of an actual market transaction. It is the scale on which such valuations would be required for a wealth tax that is unprecedented. Their survey of these existing valuation methods highlights several conditions that are imposed on the hypothetical transaction: for example, that there is a willing buyer, no compulsion to sell, and (effectively) no transaction costs. These and other conditions help to make the valuation exercise more tractable, but mean that in practice open market value pulls apart from the value of the consumption that an individual could actually fund with the asset. Whether this is a matter for concern in the design of a wealth tax depends on what one thinks is the purpose of the tax, on which the authors do not directly express a view. The empirical analysis by Daly et al. (2021) provides initial estimates of the costs of valuation to taxpayers, under certain assumptions about when taxpayers may be required to obtain professional valuations at their own expense. The exercise is restricted to direct ‘out-of-pocket’ expenses and does not attempt to estimate the time-cost or hassle to taxpayers of arranging valuations or filing their return. The lack of data on how professional valuation costs vary by asset value and type also means that the analysis must rely on some informed guesses, although these are aided by inputs from Burgherr (2021) – also in this special issue – who draws on international experience and a small survey of professional services firms. Still, these estimates are useful in providing a starting point that can be refined using additional data once available. Loutzenhiser and Mann (2021) assess the scale of liquidity constraints under a UK wealth tax and discuss policies that could alleviate these concerns. Overall, their conclusions are quite upbeat (at least compared with the challenges posed by valuation): the number of taxpayers who would genuinely be liquidity-constrained in paying a wealth tax is low and there are available design solutions. A deeper question is whether a wealth tax should, in principle, include any dispensations for the asset rich, cash poor; and whether, in practice, such measures would enable taxpayers to manipulate their liability in ways that undermine the efficiency and fairness of the tax. A further issue, from which Loutzenhiser and Mann prescind, is again the collision with politics: even if genuine instances of illiquidity are few and far between, a wealth tax would need to withstand the kinds of emotive anecdote that have featured prominently in debates about inheritance tax and council tax. If, despite these challenges, a wealth tax were implemented in the UK, how much revenue could it raise? A robust revenue estimate requires three key inputs, each of which is addressed in this special issue. Advani, Bangham and Leslie (2021) study the scale and composition of wealth owned by UK households, with particular focus on the top of the distribution. Advani and (2021) the existing international evidence on behavioural responses to a of taxes on the in to estimate to a wealth tax. Burgherr (2021) estimates the costs that would be by under an annual wealth tax, as well as the filing costs for individual on these inputs, Advani, and Hughson (2021) the revenue that could be raised from a UK wealth tax at different and into behavioural responses and costs. main themes from these empirical First, the revenue from a wealth tax depends to a very on design This is obviously in of and and the of the tax because these revenue on a However, Advani and (2021) also draw attention to the major that design in behavioural responses. Second, as further estimates of the wealth tax revenue that could be from those at the very top of the wealth are to costs on and the constraints on the and for a wealth tax, because the of many of these costs mean that they a significant of the raised when and are set at See Advani et al. for an attempt to this and set This special issue the current ‘state of the art’ on wealth both principles and practice. The papers draw on a range of international experience and evidence, and they the latest thinking from economic theory. from a several of the papers also provide practical insights to the actual of wealth taxes, with a focus on the UK context. Despite the and of these there several in which further work – and additional data – would be The first concerns the measurement of top wealth. many the UK has no data on the stock of Advani, Bangham and Leslie (2021) rely on survey data from the for Wealth and 2020. However, face difficulties with and at the top of the 2020. and the at the very top for the UK) is small to draw conclusions on scale of wealth, Advani, Bangham and Leslie (2021) attempt an for top wealth based on public data the and the of a distribution. However, as these authors there and concerns over of the rich In the absence of a wealth tax from which to obtain data on asset one – as recommended by the Wealth Tax Commission in its final report – would be for to a new for top wealth into their main on wealth. This is the approach already in and the See et al. and approach – the of additional survey data – would be to develop new estimates using the for Saez and and 2020. with survey and data assets such as for further again to those at the very top of the wealth concerns our of behavioural responses. In do the over the general by existing also hold at the very the are over a that the but also to those with of wealth. would that those in the are more and have more to in avoidance, but there are very few studies that focus on these top it would be useful to whether certain types of – at similar of wealth – are more to be either in of their or as a result of the composition of their wealth. The is not only with data to the measurement of top wealth. It is also the case that few focus on the very more general policy do not with for those at the A for such research in the UK concerns the of to the for individuals with additional empirical evidence could be to on the question – which divided the Meade Report and Mirrlees Review – as to whether and how wealth provides additional to the the consumption that it This raises both and Adam and Miller (2021) help to in principle, we should be to – or in other words how to additional of holding wealth from that are a direct of There remains further work to how these might be in practice. – of direct to a wealth tax – it would useful to whether these additional scale directly with wealth or whether they are only at high of wealth. The authors from the and through at and a COVID-19 and from the International Institute COVID-19