In this post James Dempsey and Tom Sorell, who are working on the ethics stream of the FinCris project, comment on some of the themes that have emerged from the work of the taxation stream of the project, led by Andy Mullineux.
1. Taxation and the Financial Crisis
The financial crisis has raised serious concerns about the current organisation and operation of the banking sector in the UK. The tax regime that is imposed upon banks and other financial institutions is one place to look for failures that precipitated or exacerbated the crisis, and for tools to improve the functioning of the financial sector in the future.
There are many considerations that will influence the design of a tax regime. Here we offer some that are derived from the need to address significant ethical failings in the financial system exposed by the financial crisis. Popular narratives of the crisis highlight some particular areas of concern. One is the level of compensation awarded to employees in the sector. A related concern has been the kinds of activities in which financial institutions have engaged to generate outsized profits and rewards for their employees. In particular, the idea that they have not only engaged in ‘casino banking’ but they have also ‘gambled with other people’s money’, taking risks on the basis that they will reap the rewards if the gamble pays off, but will not bear the costs if it does not. A third concern has been that banks and their employees avoid paying the taxes that they owe. This is a specific case of the broader issue concerning the extent to which large corporations in general, not just those in the financial services sector, are able to avoid paying taxes by structuring their international operations across various national tax regimes. However, it is particularly relevant to those financial institutions which benefit from being implicitly insured by tax payers. Such institutions have been called ‘too big to fail’ (TBTF).
Although not always fully unpacked in popular narratives, these concerns identify genuine ethical issues raised by the financial sector during the financial crisis. Some current features of the tax regime exacerbate these problems. But reform of the tax system also presents opportunities to make positive contributions to tackling them.
2. Ethical Problems in the Financial Sector
Objections to the level of compensation awarded to employees in the financial sector typically do not distinguish between two ways in which arrangements may be ethically objectionable, although the notion of desert is relevant to both. The first concerns the degree to which individual rewards are sensitive to appropriate measures of performance, assuming that the best system is broadly meritocratic. One practice that has been criticised on this basis is the payment of bonuses based on revenues which are realised up front, but that do not take into account costs which may accrue in the future as a result of the same transactions. An extension of this worry concerns the question of what constitutes a ‘good’ result for which reward is deserved. Some activities may generate profit for the financial institution at the cost of being ‘socially useless’ or even detrimental. The second way in which pay awards may be objectionable is if the absolute level of those awards is too high, even if within that range what any individual receives is sensitive to an appropriate measure of performance. In such a situation individuals may argue that their relative position in a pay scale is justified by the contribution they made. However, this line of argument cannot justify the absolute values that the scale employs.
In order to determine whether the absolute values of the scale of pay in financial services are justified a number of factors will be relevant. The fist factor is merit. Just as the reward to an individual should reflect their contribution relative to that of others, the rewards that accrue to a sector should reflect its productivity relative to other sectors. If the financial sector experiences competition failures, it may be able to generate revenues out of line with its contribution. A second factor, however, suggests a limit to meritocracy – it may be possible to identify absolute levels of reward that can never be deserved, irrespective of contribution, specified in terms of purchasing power. This would be the high end equivalent of the ‘dollar a day’ measure of extreme poverty. Just as no-one should live in extreme poverty irrespective of any other factors it is plausible that no-one should control extreme levels of wealth no matter how much they have contributed.
Evidence shows that, whatever the underlying reasons, rewards in the financial sector are significantly higher than other industries. Egger et al. find a 43% premium in directors’ pay in the financial sector compared to other industries, although there is a marked variation by sub-sector, with directors of commercial banks earning significantly less than their counterparts in insurance companies, real estate (primarily real estate investment trusts), and ‘other finance’ (including investment banks, brokerage, and asset management). Egger et al. therefore conclude that the overall premium for financial services is driven by the more specialised units. Moreover, the crisis itself provided a good test of the responsiveness of rewards to performance – both with respect to the sector as a whole, and individuals within it. The results of this test were not encouraging. Cuomo describes the crisis as ‘a virtual laboratory in which to test the hypothesis that compensation in the financial industry was performance based’ and, quoting figures for the nine banks that originally received US government Troubled Asset Relief Programme (TARP) funding concludes that ‘when the banks did well, their employees were paid well. When the banks did poorly, their employees were paid well. And when the banks did very poorly, they were bailed out by taxpayers and their employees were still paid well’. One role of a tax system is to eliminate distortions in the effective operation of the market; another is to be part of a system of financial wealth redistribution in line with government policies. On both counts unfairness in the distribution of income generated by outsized bonuses is an outcome the tax system can be expected to counterbalance, if not eliminate entirely.
The lack of fairness with which the UK tax system deals with the high incomes of bankers is exacerbated by the way in which large financial institutions are able to avoid paying the taxes that governments do impose, and other taxes that might be thought to be worth imposing in the interests of fairness. Such avoidance can take two forms. It can be the result of carefully designed accounting procedures that make some business income ‘stateless’ for the purposes of taxation. It can also result from what amount to threats to the host government: “do not increase taxes on us, or we will move our business to another country!”. The second way of avoiding tax has been prominent in the UK where such threats have been issued explicitly by financial institutions, and where tax revenue from the City of London forms a significant part of government revenue. The first kind of avoidance, on the other hand, is a general problem that has been associated recently with criticism of Amazon, Google, and Starbucks amongst others. Yet it is particularly relevant in connecting the financial sector to the crisis. Financial innovations that were central to causing the crisis, specifically the creation of ‘structured investment vehicles’ (SIVs) that were held off banks’ balance sheets and facilitated the process of securitising loans, were only as profitable as they were because they facilitated such tax avoidance, and because they allowed banks to evade regulatory requirements to hold capital against assets.
The connection between tax avoidance and the financial crisis illustrates that the unfairness of avoidance goes beyond the unequal distributions of income that accrue to individuals post tax. Not only did tax avoidance in the financial sector benefit the institutions and individuals in the sector, but it also imposed significant costs on society by contributing to the crisis. To make matters even worse, one of the major costs of the crisis came from the taxpayer funded bailout of failing institutions. In the UK Northern Rock was nationalised, and the UK government still owns 84% of the Royal Bank of Scotland and 43% of Lloyds TSB. In short, the financial sector benefited from not paying tax, and then benefitted from being rescued from the consequences of this tax avoidance, a rescue funded by the rest of society that did pay tax.
Unfair remuneration practices and tax avoidance are undesirable activities that design of the tax system should address. The third area of concern mentioned above is that it has become standard practice more generally for the financial sector to engage in activities that are ‘socially useless’ or even socially detrimental. The basis of the ‘gambling with other people’s money’ criticism can be explained as the tendency for the sector to employ very high levels of leverage, in other words take on very high levels of debt, and then use these funds to finance activities such as creating, holding and trading derivatives. Amongst other things, this allows an individual or institution to increase their exposure to market movements. An increase of leverage in the financial system means that potential financial returns to leveraged institutions increase, but so do the potential losses. However, while returns can be extracted in the form of bonuses, losses accrue elsewhere. In the case that the institution is bankrupted, shareholders and debt holders lose out. But where the institution benefits from implicit insurance by tax payers, it is those tax payers who foot the bill for averting bankruptcy by providing a ‘bail out’.
The increase in leverage of financial institutions in the build up to the crisis was facilitated by two markets in which they could borrow money – the wholesale money market and the market for securitised debt. As well as the specific risks it generated, high leverage in the financial sector undermined the stability of the financial system generally. This is also socially detrimental as financial stability can be considered a good that is of value to society in general. ‘Gambling’ with borrowed money was not the only undesirable activity that became common in the financial sector. Lord Turner’s original comments about ‘socially useless’ activity referred to high volumes of highly esoteric trades that do not obviously add to the real world benefits provided by the sector’s role in financial intermediation. The explosion in the market for securitised debt is a good example of this. Not only is such activity socially useless but it has proved to be socially detrimental. Its huge complexity resulted in increased opacity as market participants were unable to understand fully the risks they were taking on, or the risks to which other market participants were exposed. It contributed to the market destabilising panic that precipitated the crisis when losses began to accrue. It also contributed to the disproportionally high rewards to market participants who received bonuses according to the value of transactions they undertook. This last outcome is connected to accusations of ‘churning’ where market participants aim to generate a high volume of transactions to maximise fee income irrespective of whether these transactions are beneficial for clients or not. This market activity is an example of the kind of undesirable behaviour that the tax system might be expected to discourage.
3. Taxation as a Response to Ethical Failure
The ethical issues discussed present a number of distinct problems for the tax system to address. The first, in the name of fairness, is to correct the distortions to the effective operation of the market that allow outsized remuneration to be unconnected to what individuals deserve – either relative to each other within the financial sector, or as a sector compared to other parts of the economy. Fairness may also require some redistribution even when the market operates effectively, in line with considerations of the unjustifiability of ‘extreme wealth’. The second challenge is to tackle tax avoidance, while the third is to address other undesirable activities – taking high risks with borrowed money, and generating high volumes of complex trades to produce fee income. An ideal response to these challenges would target the outcomes or the activities that generate them as directly as possible, so sending an explicit message about the values that are being pursued through taxation. In order to achieve these aims, various different approaches to taxation are possible. In general, such taxes may be thought of as achieving their goals through either through the generation of revenue or through their effect on behaviour, or a combination of the two.
In respect to the challenges outlined, the fairness concerns are likely best addressed by revenue raising measures that take money out of the system where it has accrued due to the inefficient operation of markets, where these include the labour market and also the various markets for financial services. This money can then be used to correct other distortions, such as replacing tax payer funding of future measures to ensure the stability of the financial system. To the extent that these measures are properly targeted at those institutions where remuneration is least responsive to the markets forces that judge performance, and that pose the greatest risk to financial stability, they should also have a desirable effect on behaviour. Discouragement of undesirable activities will be best achieved by measures that target those activities as precisely as possible either by introducing disincentives to behave in these ways or by removing incentives for doing so (some of which may be a result of the existing taxation regime). Tax avoidance will be reduced by designing tax rules in ways that are least susceptible to clever accounting practices and, perhaps more importantly, by international cooperation to prevent avoidance that relies on transfers of funds across international borders or on threats of leaving one jurisdiction for another.
A number of proposals have been made for reforms to the way financial institutions are taxed in the UK in response to these problems. At a European level there has been support for the introduction of a ‘financial transactions tax’ (FTT), which is a very broad based tax to be levied on all financial transactions or the value of those transactions. The primary motivation for introducing such a tax appears to be the potential it has for revenue generation. However, such a tax would not fit very well with the requirements outlined above since it would not be a tax either on excessive remuneration or contribution to systemic threats from TBTF banks’ risk-taking. Given that these are the two major ethical rationales for raising tax revenue from the financial sector, the FTT would appear to miss its target. This conclusion is broadly in line with that of the IMF. The FTT would also not be particularly well suited to achieving behavioural aims. While it would discourage transactions generally by raising their cost, it would not necessarily reduce the attractions of socially useless or detrimental transactions specifically, relative to those that are desirable. An alternative would be a more targeted tax on particular transactions considered undesirable. This would be in the spirit of the original ‘Tobin tax’ which was proposed to be applied to currency transactions in order to discourage currency speculation and improve stability of exchange rates. Insofar as an FTT could be used to address any of the challenges highlighted, a targeted version looks to be the most promising.
An alternative to an FTT is a ‘financial activities tax’ (FAT). Rather than taxing individual transactions, a FAT targets the total value added by a bank, as represented by its profits and bonus pool. While this tax is just as blunt an instrument as FTT for discouraging particular kinds of transaction, it has the benefit of allowing bonus pools to be taxed directly – a primary aim of a revenue raising tax. Indeed, FAT forms one part of the IMF’s recommendations for new revenue raising mechanisms in the financial sector. The other part of their recommendations is a ‘financial stability contribution’ (FSC), which would be adjusted to take account of an institution’s riskiness and contribution to systemic risk. This contribution would then be put towards the future cost of ensuring systemic stability. A combination of and FSC supplemented by an FAT addresses well the two ethical issues highlighted that justify imposing additional revenue raising measures on the financial sector, and for this reason the IMF’s proposals should be favoured over an FTT. The latter should be employed, if necessary, to target very specifically the kinds of transactions that are socially undesirable.
A related proposal concerns the imposition of value added tax (VAT) on financial services. In Europe banks are exempt from VAT, which is imposed on the value of a business’ output, and removing this exemption would in principle mean that every time a bank sold a product or service they would have to add tax (currently 20% in the UK) to the price charged. As with the FAT and broad based FTT, an imposition of VAT could reduce the value of engaging in financial transactions generally, but would not target specific transactions that might be judged undesirable. The difference between VAT and an FTT is that the former is intended as a tax on consumption as opposed to transactions. There are various reasons why removing the exemption of financial services from VAT may be appropriate. However, insofar as it does not tax excessive remuneration or contribution to systemic instability, or provide specific disincentives for the kinds of undesirable activities already identified, VAT will not address the ethical issues in the financial sector identified here. FAT is sometimes seen as an alternative to VAT insofar as it taxes the total value added by the sector at an institution level, rather than on a transaction by transaction basis, and FAT has been suggested as an interim measure until an extension of VAT is possible. However, given the advantages of FAT from an ethical perspective – it targets explicitly the outsized profits and bonus pools of the sector – it is, at least in this respect, the better alternative.
An alternative proposal focused on influencing the behaviour of financial institutions does not address the volume and nature of financial transactions directly, but rather the incentives that those institutions have for funding their activities by borrowing (i.e. through debt) as opposed to through capital provided by the business’ owners (i.e. through equity funding), and hence running high levels of leverage. At present businesses are allowed to treat interest payments on debt as a business expense, and hence deduct the cost of these payments from taxable income. Costs of servicing equity funding through dividend payments are not deductible in this way, which gives businesses a reason to favour debt. The Mirrlees report offered three alternative approaches to removing this incentive to fund activities with debt: the introduction of a tax on company cash flow rather than income or profits; making the costs of equity funding tax deductible in the same way as interest payments through an ‘allowance for corporate equity’ (ACE); and removing the tax deductibility of interest payments through a ‘comprehensive business income tax’ (CBIT). The first two options effectively allow the equal treatment of debt and equity by extending the tax deductibility of funding debt to the funding costs of equity as well. The third option removes the tax deductibility of debt. While the Mirrlees report highlights problems with the third option, particularly for financial institutions, removing the tax exemption on interest payments would be preferable from an ethical perspective since it targets the badness of this exemption explicitly, rather than trying to mitigate it through an alternative treatment of equity. It could even be that such exemptions would only be removed for financial firms, emphasising their central role in creating dangerous leverage in the economy.
These suggestions for addressing market distortions associated with excessive compensation and contribution to systemic risk, and for discouraging socially undesirable activities, still leave open the question of how the UK, or any other country, can implement them successfully. Any implementation of a new taxation regime should also address implementation failures of the existing system, particularly the possibility of ‘stateless’ income. Kleinbard argues that the imposition of ‘worldwide’ tax systems – where governments tax companies headquartered in their country on all the income they generate worldwide – would be the best option for addressing this challenge. This would not, however, solve the problem of financial organisations which threaten to relocate their headquarters. Reforms to bank taxation will only ultimately be partially successful as long as there is a lack of international coordination in their application.
The major ethical issues that have arisen out of the financial crisis, and that have been prominent in the calls for action that have followed, have been several fold. One is the perceived lack of fairness in the way that governments treat financial institutions. Employees of such institutions are allowed to reap huge rewards seemingly unconnected to any measure of what they deserve; and the institutions and their employees avoid paying taxes, while at the same time benefitting from implicit insurance funded by tax payers. Another major ethical issue is the degree to which financial institutions undertake activities that are harmful to society, such as establishing risky, highly leveraged business models on the basis that when things go well rewards will be enjoyed privately, but when things go badly costs will be socialised. A final issue is the degree to which financial institutions are perceived to have governments over a barrel, threatening to jump ship to another jurisdiction if actions are taken which threaten their ability to make money.
Taxation offers some tools for addressing these issues. A financial stability contribution and financial activities tax target fairness concerns directly by being imposed on the basis of the degree of government support the institution has needed or will need, and on the size of profits and bonus payments. The solution to penalising socially detrimental activities will be less direct, if the favoured approach is to try to remove an imbalance in the incentives for debt and equity finance to reduce the tendency to leverage in the financial sector. While this move is important, a more direct approach to addressing undesirable activity would be preferable – for example by singling out certain kinds of transaction for special taxes, whether they are short term wholesale money market transactions, or esoteric derivatives trades. Any of these solutions will need to tackle ways in which the financial sector will resist their implementation by playing one country’s taxation regime off against another. Here, there will be no substitute for international cooperation.
 Indeed, the thought that the financial sector does experience fundamental breakdowns in competition has been a popular theme since the financial crisis.
 ‘Commercial banks’ are those whose business is focused on taking deposits and issuing loans, particularly to business customers, as opposed to providing in more specialised or exotic financial services.
 Egger, P. H., Ehrlich; M. V, Radulescu, D (2012). How Much it Pays to Work in the Financial Sector. CESifo Economic Studies 58(1): 110‐139.
 Cuomo, A. M. (2009). No rhyme or reason: The ‘heads I win, tails you lose’ bank bonus culture. Attorney General State of New York. July, 30.
 The ‘too big to fail’ phenomenon is a market distortion since the potential failure of such an institution creates an ‘externality’, a cost that is not borne by the institution itself or its members (either through its collapse, or its rescue by tax payers) and hence not reflected in its decision making. Taxation, or regulation, can correct this problem by forcing the institution to ‘internalise’ the costs of insuring itself against failure.
 A general aim of wealth distribution could be attained by economy wide measures such as changes to taxes on income and wealth. There may well be grounds for such measures. However, they would not be sensitive to income and wealth distorting features of particular sectors of the economy, such as the financial sector.
For a detailed treatment of ‘stateless income’ see Edward D Kleinbard, ‘Stateless Income’s Challenge to Tax Policy Stateless Income’s Challenge to Tax Policy’, Tax Notes
 James Moore, ‘HSBC in New Threat to Leave the UK over Osborne Banking Levy’, The Independent, 6 November 2010.
 ‘The Top Poacher’, The Guardian, 6 February 2009.
 Given that this regulatory requirement is equivalent in many respects to a tax on banks, its evasion can be considered another incidence of tax avoidance.
The promise of such government action to save failing banks was never explicit, but there was an implicit acknowledgment that an institution would not be allowed to fail if it was ‘systemically important’. What constitutes ‘systemic importance’, however, is itself unclear and in a time of crisis it may be though that any failure of a financial institution would risk catastrophic knock-on effects. Moves towards ‘living wills’ and the introduction of bank levies or ‘financial stability contributions’ (FSCs) adjusted by an institution’s risk profile are intended to address such issues (see section 3).
 While the bail-out itself was not directly a matter of tax policy, the fact that it was needed at all resulted from a failure of tax and regulatory policy to deal with TBTF banks.
 In this case derivatives were used to increase exposure to risk. They can equally be used to hedge risks, so increasing safety but also reducing potential returns. One of the difficulties in controlling risky activity is to determine the purpose of any particular trade and limiting only the undesirable ones.
For example, if an individual has £5 and buys £5 worth of shares with it, a 10% increase in the share price will yield her a 50p profit. If she instead buys £20 of shares, putting her £5 down as a deposit and borrowing the remaining £15, a 10% increase in the share price will yield her a £2 profit on her underlying equity of £5. Of course, this magnifying effect works with losses as well as profits.
Wholesale funding involves financial institutions securing funding from other financial institutions, non-financial corporations, foreign entities and state and local authorities, typically on a short term basis (See R Huang and Lev Ratnovski, ‘The Dark Side of Bank Wholesale Funding’, Journal of Financial Intermediation, 2011, 1–46. The market for securitised debt, on the other hand, allows financial institutions to package the cash flow from loans they have issued in the form of securities, and sell these securities to third party investors, essentially borrowing money from them for the duration of the investment
For more on financial stability as a public good, see Andrew W. Mullineux, ‘Taxing Banks Fairly’, International Review of Financial Analysis, 25 (2012), 154–158: 2-3.
 Adair Turner, ‘How to Tame Global Finance’, Prospect Magazine, 2009.
 To be clear, securitisation is not necessarily an undesirable practice. Indeed, it can provide an important alternative mechanism for stimulating the flow of money to consumers and businesses, especially when banks are reluctant to lend. However, it was the size and complexity of the market in securitised debt and associated products (such as credit default swaps) that led to Turner’s accusation of socially useless activity.
 IMF, A Fair and Substantial Contribution by the Financial Sector, 2010.
 J Tobin, On the Efficiency of the Financial System, 1984.
 In an ideal world an FSC that decreases with capital held is equivalent to imposing capital requirements through regulation. However, given that the ‘buffers’ lie in different places under the two options – the government in the former, and the institutions in the latter – the practical results of the two are not the same.
In practice the tax levied on that product or service would not increase from 0% to 20% as the bank would be able to claim back VAT that it had paid on the inputs it had used.
 See Agarwal, Chaudhry, and Mullineux, ‘Bank Taxation and Regulation’, unpublished.
 IFS, Tax by Design, 2010.
 ‘Taxing interest income while giving no tax relief for interest payments would imply a huge tax increase for banks and other intermediaries that make profits from borrowing and lending. Conversely, a symmetric treatment of interest income and interest payments would imply no taxation for interest income. This would exempt from taxation the component of bank profits that results from interest spreads. Neither of these outcomes seems to be attractive. One alternative would be to make interest payments deductible only against interest income’ IFS, op. cit.
Kleinbard, op. cit.