The UK's Exit Charge from the EU: Insights from Modes of Accounting

Whatever the final charge on the UK for leaving the EU, the money itself is relatively marginal to the former's public finances. However, this charge is politically sensitive and financially aggravating during one of the longest periods of fiscal austerity in the UK's history. The ways in which leaving is conceptualized have implications for any continuing financial obligations that must be managed within the context of fiscal austerity and political uncertainty. Yet, leaving the EU is a unique transaction: it is not analogous, for example, to a divorce settlement, the leaving of a club, the termination of a commercial contract, the leaving of a treaty&#8208;based international organization, or secession from a state. Analyzing the formulation of the charge in terms of the four modes of government accounting&#8212;financial reporting, statistical accounting, budgeting, and fiscal sustainability projections&#8212;enhances its fiscal transparency. It evidences not only the weakness and inconsistency of the UK's negotiating position but also the dominance in EU thinking of the short&#8208;term budgetary calculations of the 2014&#8211;20 Multiannual Financial Framework over its long&#8208;term sustainability without a large net contributor. The final amount paid by the UK will depend on the resolution of competing perspectives as well as on liabilities and contingent liabilities associated with the increasingly complex EU financial architecture.

Nor do I remember any serious opposition to the idea, enshrined in the Lisbon Treaty in what became Article 50, that nation-states were entitled to change their minds, and leave if they so choose. Equally I'm certain no-one dreamed that in 2017 a member state would trigger the procedure, as Mrs. May did on 29 March. However, once Article 50 had been triggered, the UK's status changed to that of a third party, governed by Article 218(3), and certain UK citizens working in EU institutions ceased to have access to documentation and decisions on the basis of their allegedly conflicted loyalties. Control of the exit process passed to the European Commission, acting on behalf of the Council of Ministers (i.e., the governments of the Member States), now the 'EU27'. Pre-conditions were set that the Commission would not discuss the UK's future trading relationship with the EU until three issues were settled: the treatment of non-UK EU citizens resident in the UK and of UK citizens resident in the EU27; the arrangements for the border between Northern Ireland and the Republic of Ireland; and the exit charge that the UK would pay.
The UK net contribution to the EU is £8-10 billion per annum (ICAEW, 2017), approximately 1% of annual UK public expenditure. A much-cited early figure for the exit charge was €60 billion if leaving on 29 March 2019, which if added to the UK public sector net debt would increase the net debt/GDP ratio from 87% to 90%. Before the global financial recession, the 2007 ratio was 36% (ONS, 2018). The draft Withdrawal Agreement (European Commission, 2018) has been interpreted by the UK's fiscal council as involving an undiscounted payment of €41.4 billion, assuming a transition period ending on 31 December 2020 (OBR, 2018, para. B.35). Such a charge is small compared with the potential effects of Brexit on GDP growth and on UK public finances, for example due to sterling depreciation and trade disruption. The Institute for Fiscal Studies estimated the annual budgetary cost as £70 billion (Emmerson et al., 2016).
Quickly labelled the 'Divorce Bill', the exit charge generated considerable political acrimony, despite its relatively marginal economic impact (Keep, 2017(Keep, , revised 2018: If you were sitting in a bar and if you are ordering 28 beers and then suddenly some of your colleagues is leaving and is not paying, that is not feasible. They have to pay-they have to pay. Not in an impossible way, I am not in a revenge mood. I am not hating British. The Europeans have to be grateful for so many things Britain has brought to Europe during war after war, before and everywhere and every time. But now they have to pay (Jean-Claude Juncker, President of the European Commission, 2017). The sums that I have seen that [the EU27] propose to demand from this country seem to me to be extortionate. I think that to "go whistle" is an entirely appropriate expression (Boris Johnson MP, then UK Foreign Secretary, 2017). I am not hearing any whistling, just a clock ticking (Michel Barnier, EU Brexit Negotiator, 2017).
Handing over an allegedly huge sum to Brussels while the British people could see no end to austerity at home gave the exit charge a high political salience. Moreover, the EU27 insisted on an early financial settlement before negotiations could move on to trade issues vital to the UK (European Council, 2017). The UK Government thought that it could deal bilaterally with Member States, by-passing the European Commission, a tactic which underestimated the cohesion generated by survival through the eurozone crisis.
Counterfactual history is always problematic, but there is widespread academic and commentator agreement that post-2008 fiscal austerity was a significant factor in bringing about the Brexit result. What was markedly different about the UK Government reaction to the post-2008 fiscal crisis was that it focused almost exclusively on expenditure cuts rather than the usual mix of spending cuts and tax increases (Mauro, 2011). Austerity has also lasted much longer than the customary two or three years of fiscal pain, followed by relaxation, that has characterized UK fiscal history over the last 100 years (Hood and Himaz, 2017). Stuckler et al. (2017) report that austerity has hurt deprived groups through a number of channels, such as unemployment, poverty, homelessness, and reduced access to healthcare. Stiglitz (2011) claimed that 'draconian cuts are causing many [UK] people to lose hope'. The projected Brexit dividend of more money (£350 million per week) 2 for the National Health Service was therefore particularly attractive to disadvantaged groups. Sending money annually to Brussels had become toxic, thereby increasing the sensitivity attached to the exit charge. Remarkably, the eventuality of such a charge had not featured in the Brexit Referendum campaign, suggesting a blind spot on both sides about the nature of the UK's relationship with the EU.
Without the 2010-2016 period of post-crisis UK fiscal austerity, there might not have been a Brexit majority, even if internal Conservative Party calculations had led then Prime Minister David Cameron to call an In-Out Referendum. Moreover, without the 2008 global financial crisis and the 2009 eurozone crisis, the EU's development path would have been different, thereby avoiding the further disengagement of the UK (Laffan and Schlosser, 2016). Brexit added another element to the crises already faced by the EU (Laffan, 2016), notably the 2009 eurozone crisis, instability on its Eastern border, and migration flows from failed states in the Middle East and North Africa. One segment of 'Brexiter' opinion considers this an opportunity to complete those parts of the state-shrinking Thatcher revolution that had been frustrated by EU membership (Lawson, 2016). 2 The Institute for Fiscal Studies (Emmerson et al., 2016, p. 2) disputed the extra £350 million a week the Brexit Leave campaign claimed would be available for spending on the NHS after Brexit. This figure was calculated without the UK's receipts from the EU and the Fontainebleau Abatement being taken into account: the correct figure was £150 million a week, calculated on the assumption that Brexit would have no other effect on UK public finances. Yet the number was widely believed and has since been repeated by Boris Johnson, the then UK Foreign Secretary, leading to a rebuke from the Chairman of the Statistics Authority (Norgrove, 2017). The Fontainebleau Abatement derives from a political deal won by UK Prime Minister Margaret Thatcher in 1984 whereby a reduction is applied to the UK gross contribution before money is transferred to the EU. Otherwise, the UK would be a much larger net contributor to the EU budget (ONS, 2017). This aspiration has made the EU27 nervous about its future relationship with the UK, fearing a tax and regulatory race to the bottom. Our first research objective in investigating the UK's exit charge from the EU is to treat the calculation of the exit charge as a natural experiment and illuminate it using the four modes of government accounting: financial reporting, statistical accounting, budgeting and fiscal sustainability projections (Heald and Hodges, 2015). This analysis draws on public documentation and calculations that neither governments nor supranational organizations would normally reveal. Our second objective is to take the exit charge as an unusual opportunity for demonstrating the value of using the four modes of government accounting as an investigative method, particularly in the current context of austerity.
We explore alternative conceptualizations of the UK's break from the EU, noting that adherence to different conceptualizations contributes to conflict in Brexit negotiations. The four modes of government accounting provide the foundation for our analysis. While we draw on our prior involvement in UK and EU public sector accounting developments, we have enjoyed no insider access and have relied on documents in the public domain. This has been less of a disadvantage than it might seem, because both the EU and UK sides have extensively leaked to the media their version of the rights and wrongs of the exit charge. Thus data have become available that EU institutions and governments would either not calculate (on grounds of hypotheticality) or refuse to release (on grounds of 'national' or 'supranational' interest). We have tracked events, with the Financial Times being particularly useful; drawn on the analysis by the Office for Budget Responsibility (OBR, 2018) of the January 2018draft Withdrawal Agreement (European Commission, 2018; and have participated in seminars held under the Chatham House Rule 3 which facilitated contextualization and interpretation. All have contributed to our understanding of why the Brexit exit charge has become so conflictual.

CONFLICTING CONCEPTUALIZATIONS OF BREXIT
As in marital divorces, both the UK and the EU secured legal advice confirming their own position in the financial dispute. A report by the House of Lords European Union Committee (2017, para. 137, p. 39) concluded that, in the absence of a withdrawal agreement, the UK has no legal obligation to pay to exit, but that the 'political and economic consequences … [of not paying] … are likely to be profound'. No international court would have jurisdiction. The European Court of Justice (ECJ) would not have jurisdiction because the UK would not be a Member State. Because the EU is not a state, the International Court of Justice 3 Named after the London headquarters of the Royal Institute of International Affairs, the Chatham House Rule governs the conduct of policy seminars at which the identity of speakers and participants is never made public, but the substance of discussion can be reported on a nonattributable basis. In particular, this protects civil servants and others in exposed positions.
would only have jurisdiction if the UK and the other 27 Member States had made a declaration submitting to its jurisdiction, and they would not have done so. If the UK were to refuse to pay its liabilities, the EU would not conclude future trading and other agreements with the UK. Though evocative shorthand, the divorce conceptualization of Brexit is only one of those articulated. Others include quitting a club, terminating a contractual relationship under private law, leaving a treaty-based international organization, and seceding from a state. Of central importance underlying them is the UK's traditional transactional approach to EU membership, always assessing costs and benefits. This has applied across the UK political spectrum and has characterized all UK governments since entry into (what became) the EU in 1973. The latearriving UK never embraced the existential 'peace and prosperity' vision of the EU that was shared by founder members. Rogers (2017) traced the origins of Brexit to the 1992 Maastricht UK opt-outs on the single currency and Schengen border control, and particularly to the 2011 UK veto of treaty changes desired for purposes of fiscal co-ordination by the eurozone countries at the height of their fiscal crisis. It is not that, for example, France and Germany do not themselves pursue self-interest in economic and fiscal matters, but they do share a European vision to which the UK has never subscribed.
There is irony in that successive UK governments pressed for early membership for the former communist states in Eastern Europe, with the purpose of diluting ambitions for political integration on the lines of the 'ever closer union' expressed as a political goal in EU Treaties (Miller, 2015). Although these countries can behave as transactionally as the UK and therefore should be natural allies, resentment of Eastern European immigration into the UK was a powerful factor in the Leave campaign, in turn undermining the potential for Eastern European states to be UK allies.
The competing conceptualizations are summarized in Table 1, which sets out the justification, the appeal to policy actors, and the financial implications if that conceptualization were adopted.

Brexit as Marital Divorce
Marital divorces are complicated and how they are constructed has undergone significant legal change, particularly affecting financial settlements. The English courts have favoured 50:50 splits of net assets, irrespective of wealth taken into a marriage, relative earnings during the marriage, and projected future earning power after the divorce. If Brexit were a marital divorce then, on this basis, the UK would receive back its share of net assets or pay over its share of net liabilities at the settlement date. The notion that the exit charge would be calculated on net assets or net liabilities was rendered implausible by the timing imposed on the UK by the EU27, once Article 50 had been activated. Although Barker (2017) did calculations based upon the European Commission's financial accounts, the EU27 has had no intention of letting the UK take away a share of the EU's net assets. Disruption having been caused by the UK's decision to leave, it must pay its share of financial liabilities but would in general have no claim on EU assets. Certainly,  T H E U K ' S EU EXIT CHARGE 563 this will not be a divorce of the kind obtainable from the English courts at the dissolution of a marriage.

Brexit as Quitting a Membership Club
The intense UK political and media rhetoric about extortion and 'let the EU go whistle' included indications that leading Brexiters likened Brexit to leaving a club. Joining a club usually involves paying a joining fee and then an annual membership fee. 4 If the member exits the club, they would give a period of notice, and settle any outstanding fees and charges. But the departing member does not receive a share of net assets at the date of departure nor does it have to fund a share of net liabilities. The UK was perhaps never serious in claiming a share of net assets, but it would like the clean break of the club scenario. However, the major figures of the EU27 do not see the EU as analogous to a membership club; they hold departing members responsible for their share of liabilities and contingent liabilities.

Brexit as Terminating a Commercial Contract
In a commercial contractual relationship between two private entities, the relationship will only survive long term if both parties see future gain to themselves. This sense of continuous calculation is close to the UK's transactional approach, but alien to the European vision (Laffan 2016;Rogers, 2017). In the days when the large auditing firms were partnerships, one bought into the partnership at entry and was bought out at exit. Because of unlimited liability one was jointly and severally liable during the partnership but free from liability after departure.

Brexit as Leaving a Treaty-based International Organization
Countries can walk out of international organizations because of policy disagreements, with any financial payments dependent on the power of the departing state. A recent example is the US's decision to leave UNESCO in 2018, having suspended subscriptions since 2011 (UNESCO, 2017 (Treasury, 2014). In summary, two conclusions deserve emphasis. First, much argument is opportunistic, with actors calling on principles that support their desired outcome. This is no surprise but it makes satisfactory resolution more difficult when public positions harden and UK negotiators expected allegations of betrayal and sabotage from behind them. Second, conflicting understandings of the UK-EU relationship coalesce with deliberate misinterpretations of accounting and statistical data. We now turn for illumination to the four modes of government accounting.

FOUR MODES OF GOVERNMENT ACCOUNTING
Power politics dominated the fraught exit charge negotiations between the UK and the EU. On 12 June 2017, the EU27 published a statement of principles governing the calculation (European Commission, 2017a) whereas the UK consistently refused to state publicly its position, while engaging in political rhetoric. A conditional offer of circa €20 billion was made in the UK Prime Minister's Florence speech on 22 September 2017(May, 2017b, designed to tone down the toxicity that had built up after her Lancaster House speech on 17 January 2017 (May, 2017a).
Pioneering work in the 2000s by Frank Eich, who was then responsible for the UK Treasury's long-term fiscal projections, is conveniently summarized in Eich (2008). Figure 1 reproduces his conceptualization of the public sector balance sheet. This facilitates an exposition of the four modes of government accounting: financial reporting, statistical accounting, fiscal sustainability projections, and budgeting.
In countries that have led public sector accounting reform, accrual accounting has replaced variants of cash accounting and modified accruals. Figure 1 has four quadrants, the vertical dimension distinguishing between assets and liabilities and the horizontal dimension between events in the past and in the future. It illuminates the gains from having a public sector balance sheet, but also the gaps that affect-to varying degrees-both financial reporting and statistical accounting. The shaded rectangles are those included in a financial reporting balance sheet. The unshaded 'Future liabilities incurred in the future' is an important omission.
Whereas financial reporting provides comprehensive coverage of liabilities accumulated to date (the bottom left quadrant), statistical accounting generally does not include provisions that arise from past events. Both modes of accounting attach central importance to recognition criteria. For example, certain items are not recognized in balance sheets because they are executory contracts, meaning that there is no accounting recognition until delivery. Therefore, though organizations have contractual obligations to employees, future employment costs are not put in the balance sheet as liabilities. Until relatively recently, public sector organizations did not report accrued employee pensions liabilities.
Public sector balance sheets do not include future taxation revenue (top right quadrant). Of most relevance in the present context is what lies in the bottom right quadrant: a. future liabilities from past activities (which financial reporting seeks to cover comprehensively, unlike statistical accounting); and b. future liabilities incurred in the future.
The innovation of fiscal sustainability projections directs attention to (b), which fail accounting recognition criteria but which hang over future public finances. An example is the to-be-accrued pension liabilities arising from the future employment of existing and new public employees. Unlike (a), these fail accounting recognition criteria, being treated as executory contracts.
With regard to the top left quadrant, financial reports are prepared on the going-concern convention: the default assumption is that the organization will continue in broadly the same shape, irrespective of whether the measurement basis is historic cost or some form of current cost or fair value accounting. Herein lies one difficulty for exit charge calculations that seek a basis in annual financial reports. For example, the reported net assets of the European Investment Bank (EIB) are irrelevant to a calculation that includes an offset for assets. The relevant number would be the UK's share of the hypothetical flotation value of the EIB.
Fiscal sustainability analysis, taken over from the UK Treasury on the establishment of the Office for Budget Responsibility in 2010, is relevant to the exit charge. This involves forecasting cash flows over 50-year and infinite time horizons, on the basis of 'existing policies'. The calculation of fiscal gaps indicates the extent of fiscal unsustainability to be resolved by increases in taxation or reductions in expenditure. Even at the national level, there are serious difficulties in establishing in operational terms what constitutes existing government policies. The economic and demographic uncertainties are profound. What happens over time, in terms of crystallization of the contingent liabilities relating to the increasingly complex EU financial architecture, should be of profound importance to the exit charge calculation. If assumptions are made now, a lump-sum exit charge can be calculated, whether that is handed over as a single payment or in stages. Alternatively, the final amount of the exit charge will be influenced by future economic conditions and EU decision making on how to manage liabilities (such as the indexation of EU employee pensions) and contingent liabilities (such as the future willingness to write off EU loans to organizations and countries, including Member States).
Notwithstanding the focus on accounting that is emphasized by the quadrants of Figure 1, there is plenty of evidence that it is budgeting that decision makers care about, much more than the later financial reports (Public Administration and Constitutional Affairs Committee, 2017). Unlike statistical accounting (on Eurostat standards) and financial reporting (more harmonization broadly on IFRS/IPSAS standards), budgeting processes remain largely the responsibility of nation states. There are wide differences, especially on the breadth of coverage of public institutions and in the accounting basis (cash, accruals, or variants). The common feature is that executive decision making (Diamond, 2013) and the acquisition of legitimacy through legislative endorsement (Lienert, 2013) both use budgeting numbers, however those are constructed.
Of critical importance is the way in which the EU conducts its financial programming within the framework of the 2014-20 MFF. This is not a seven-year budget but facilitates the implementation of common policies and informs beneficiaries and finance ministries. The MFF follows a special acceptance procedure: proposed by the European Commission, voted on by the European Parliament on a Yes/No non-amendable basis, after which the European Council can make changes without going back to the Parliament. The MFF has been regarded as binding by recipient and contributing countries, though actual payments can be frustrated by restricting the annual budget. Unspent funds in the MFF accumulate, and are known as Reste à liquider (RAL). De Wilde (2012) found that the MFF process was characterized by intergovernmental polarization (each Member State calculating its contributions or receipts) rather than transnational polarization (interest groups coalescing across Member States).
The point to be stressed is the different ways in which the UK and EU undertake their budgeting. The UK has Spending Reviews; their periodicity, years covered, and content are under Treasury control. Spending Reviews are conducted on an accruals basis and tend to cover three years ahead; they are never voted by Parliament. Formal authorization, again on an accruals basis, takes place after the financial year has started, through what is known as the supply procedure. Unspent amounts in voted estimates expire at financial year-end, and have to be voted again, even when the Treasury has operated a carry-over system. In 2010, the incoming Coalition Government cancelled all accumulated end-year flexibility that had built up during the 1997-2010 Labour Government.
In contrast, the EU operates on a dual commitments (seven-year MFF) and payments (annual budget) basis, in which unspent commitments carry forward and do not automatically expire, though they can be decommitted. Hawkish Member State attitudes to authorizing payments in the annual budget, often with the UK in the forefront, have prevented commitments in the MFF being fully funded for individual years, leading to a build-up of unexpired commitments (i.e., RAL). 6 Working from its own practices, the UK thinks of unspent commitments on 30 March 2019 as not being its responsibility. In contrast, net recipient EU27 countries are programming that expenditure into the 2020s, considering the MFF amounts to be a binding obligation on all the EU28.

THE SUBSTANCE OF THE BREXIT CHARGE DISPUTE
In terms of headline numbers, the European Commission asked for circa €60 billion and the UK, after initially denying that it had anything to pay, made what was interpreted as an offer of €20 billion in the Prime Minister's Florence speech (May, 2017b). It did not seem coincidental that €20 billion is about two years' UK net contribution, thereby filling the budgetary hole in the final two years of the 2014-20 MFF. Sterling depreciation of 14% against the euro since the Brexit Referendum increased the sterling cost of the exit charge payable in euros (European Commission, 2017a).
Much discussion about the exit charge has centred around the EU budget and the European Commission consolidated financial report. However, there is a much broader context, as shown in Figure 2, which is a schematic representation of an official diagram (European Commission, 2017b, p. 9). Figure 2 demonstrates the increasingly complex EU architecture. The shaded circle represents the EU budget, which might be thought of as a planet. There are many moons, some intersecting with the EU budget and falling within what is known as 'full EU accountability' (represented by the outer circle). This term means that the organizations and/or funds are audited by the European Court of Auditors (ECA) and subject to discharge by the European Parliament.
On the right of Figure 2 and intersecting with the EU budget and full EU accountability is a circle representing the EIB, which is an EU institution not consolidated in the accounts of the European Commission. The area of intersection contains, for example: financial instruments and EIB shares (within the EU budget); European Fund for Strategic Investments guarantees (partly inside the EU budget and wholly within the outer circle of full EU accountability); and the European Financial Stability Mechanism and Euratom loans (outside the EU budget but inside the outer circle). Contingent liabilities sometimes overlap the budget, and sometimes do not.
On the left of Figure 2, outside both the EU budget and full EU accountability, are the institutions connected to the eurozone, notably the European Central Bank (ECB) and the European Stability Mechanism. The UK's multiple opt-outs mean that it has limited involvement in this area. Moreover, the difficulty of making treaty revisions, to which the UK has contributed, has increased the use of intergovernmental agreements between subsets of EU Member States. This is also a mechanism by which Member States, acting through the European Council, bypass the European Commission (Laffan and Schlosser, 2016).
In summary, this architecture reflects not only the growing complexity arising from the co-existence of the eurozone 19 and the non-eurozone 9, but also offbalance sheet activity on behalf of the EU28 and political competition between Member States and the European Commission. Sinn (2015) has criticized these developments as constituting a 'shadow budget' which-if not checked-will grow non-transparently alongside tight control of the EU budget.
If Brexit were analogous to a divorce on the basis of a pro rata split of net assets, there would be a valuation on Brexit day of everything in Figure 2 relating to the UK. Barker (2017) attached a total EU assets valuation of €22.5 billion, providing the UK with an offset of €2.7 billion (12% share) or €3.4 billion (15% share). 7 Alongside France, Germany, and Italy, the UK is the equal largest shareholder in the EIB, with 16.11%. Ceasing to be an EU Member State renders it legally unable to continue as a shareholder. The EU27 have no intention of allowing the UK to take core assets with it; Brexit is seen as analogous to secession, not to divorce, and discouraging imitation is a high priority.
Several complications for the financial settlement have arisen since the activation of Article 50. First, the EU view of the likely UK exit liability was first promulgated by well-briefed articles in the Financial Times (summarized in Barker, 2017), providing indicative numbers for total EU liabilities and alternative methodologies for calculating the UK share. One of Barker's calculations is summarized in Table 2. On a UK share of 15%, Barker shows Total Liabilities (€73.3 billion), and Contingent Liabilities (€11.9 billion), against which are €18.4 billion of possible offsets. That gives a total of €66.8 billion. The comparable figure on a 12% share is €51.4 billion.
In June 2017 came the official publication of the EU's principles for calculating the exit charge, though without numbers (European Commission, 2017a). The numbers reported by Barker (2017) were interpreted in the UK as an opening gambit: however, Jean-Claude Juncker, President of the European Commission, noted that the financial calculations were more complex than expected, but that the British would 'have to pay' (Juncker, 2017). In contrast, the UK Government has never published its own analysis of the UK liability, though ministers rubbished the EU figures as extortion, punishment, and ransom. It became clear that UK Prime Minister David Cameron's pre-Referendum instruction that the civil service would make no preparations for Brexit had been obeyed. The UK argument that it would accept liability only for those EU-employee pensioners who are UK nationals was clumsy (if it were a tactical ploy to have something to concede later on) or inflammatory (if serious).
Second, the UK's liability is affected by the appearance on the UK agenda of a 'transition period' after 29 March 2019, possibly of two years. During this period, the UK would be in the departure lounge: not a Member State, so having no representation, but subject to the usual budgetary contributions, all EU law (including that newly coming into force), and subject to the jurisdiction of the ECJ. Although seemingly implausible in the aftermath of the Brexit Referendum, this 'transition period' (EU terminology) or 'implementation period' (UK terminology) is included in the draft Withdrawal Agreement (European Commission, 2018). Significantly, the EU restricted duration to 21 months whereas the UK had asked for 24 months: this means that the UK effectively leaves on 31 December 2020, the final day of the 2014-20 MFF. Such an arrangement solves the short-term budgetary gap which worries both net recipient and net contributing Member States, as two more years of the 2014-20 MFF will have expired. However, the issue of unspent commitments remains: on past experience, significant amounts of RAL will continue until at least 2023, and some for much longer (European Commission, 2015). Further involvement of the ECJ and the ECA crosses 'red lines' set by the UK Government for internal party management purposes. Another issue is that impending Brexit may reduce the amount of EU receipts (e.g., from competitively tendered programmes such as Horizon 2020) and thus increase the exit charge beyond estimated amounts.
Third, threats to fiscal transparency have become evident. Having elevated the exit charge to such prominence, the pressures to conceal the amounts payable mounted. Rather than a clean break (pay the agreed financial liability as a lump sum as total discharge, then pay for participation in particular programmes), there will be staged payments. Payments that arise from treaty obligations generally fall within the accepted areas where UK payments can be classified as Consolidated Fund Standing Service, which leads to an automatic charge on the Consolidated Fund without requiring parliamentary approval. Under the European Union (Withdrawal) Act 2018, this requirement could be inserted by secondary legislation. Because of RAL from successive MFFs and the gradual crystallization of contingent liabilities, this situation could exist for a very long time (OBR, 2018).
Inadequate attention has been paid to the build-up of EU contingent liabilities, an unsurprising development after long periods of tight control over EU expenditure. There could be EU27 demands for further payments for several decades as these contingent liabilities crystallize (NAO, 2018). This crystallization process will be managed by the EU27, with the UK having no role in decisions that influence those amounts, for example, debt write-offs from the EU budget to EU institutions and third parties.

RESOLVING THE DISPUTE
The draft Withdrawal Agreement (European Commission, 2018) specifies calculation principles, without attaching numbers. Fortunately, these numbers fall within the remit of the OBR which has published estimated payments (OBR, 2018, Annex B) subsequent to an explanatory letter from the Chancellor of the Exchequer to the Chair of the Treasury Committee (Hammond, 2018). The total exit charge is estimated as €41.4 billion, roughly midway between Barker (2017) and May (2017b).
The upper part of Table 3 shows UK payments if the UK were to remain in the EU. The lower part shows UK payments with the UK leaving the EU on  Table 3 shows the annual path of payments from 2019. Although no longer a Member State from 30 March 2019, the UK would be financially treated as such until 31 December 2020. In 2023, the Leave payments will be €2.9 billion in comparison with the Remain payments of €9.7 billion. The layout of Table 3 shows how the total UK contribution depends significantly on public sector net receipts and private sector receipts. In the latest outturn year (2018), these offset 40% of the gross contribution and-without Brexit-would have been forecast to offset 47% in 2023. A fiscal risk is that a combination of continued austerity (affecting the capacity to meet co-funding requirements) and of unwillingness to engage in the context of Brexit (or being frozen out by EU27 partners) will lead to a shortfall in UK receipts.
The size of the exit charge depends on the base date from which it is calculated. If calculated from the day after Brexit day (30 March 2019), the OBR (2018, para. B.35) states it to be €41.4 billion. Using its forecast of exchange rates, this converts to £37.1 billion. If calculated from the date (1 January 2021) when the EU financial regime ceases to apply to the UK, the exit charge would be €22.9 billion. Figure 3 plots the OBR's (2018) estimates of the time profile of financial settlement payments, beginning from 30 March 2019 (i.e., after Brexit but before  (European Commission, 2018, Article 135, para 5(b)) contains a provision that the UK could then ask to settle all outstanding pension liabilities in five equal annual instalments.
Whereas Barker (2017, shown earlier as Table 2) had suggested significant numbers for the UK's share of Liabilities and Contingent Liabilities, Figure 4 suggests that the settlement of non-MFF and non-RAL liabilities involves numbers that are small in this context. Above the horizontal axis is plotted the UK's payment of its share of EU pension liabilities, running on until 2064. There is no mention in the draft Withdrawal Agreement (European Commission, 2018) of the UK paying compensation for the relocation of the European Banking Authority (Paris) or the European Medicines Agency (Amsterdam).
Below the horizontal axis is plotted the UK's receipts from reimbursements related to its stake in certain EU assets. The solid black line shows the net position, which turns negative in 2031. The most important inflow relates to repayment in 11 instalments of €0.3 billion and one of €195.9 million of the UK's paid-in capital in the EIB. This capital was contributed in years from 1973, but returns in much depreciated currency represent a fraction of the potential market value of the UK's stake of 16.11%. The UK remains liable up to its subscribed capital of €39.15 billion until the EIB's projects as at 29 March 2019 have ceased to be at risk. It is likely that other asset returns have been agreed on disadvantageous terms for the UK because current values are not being used. For example, in 2021 the UK will receive €55.51 million for its 13.6743% stake in the ECB.

CONCLUSION
This analysis of the UK's exit charge not only informs our understanding of this specific case, but also raises wider issues for public sector accounting research. Examining the issues through the lenses of the four modes of government accounting at a time of austerity brings these into focus.
The first research objective has been to use the four modes of accounting to illuminate the conflict over the calculation of the exit charge which drew heavily on financial reporting numbers. That conflict derived in part from conflicting conceptualizations of EU membership, ranging from the Brexiter notion that this was like leaving a membership club to the EU27 view that Brexit, though legally permissible under the Treaty of Lisbon, was comparable to secession. The treatment of the UK's stake in the EIB is disadvantageous to the UK, and one of the clearest indicators that this is viewed by the EU27 as secession. The analysis supports the rejection of the marital divorce, membership club, and commercial analogies, leaving only secession and quitting a treaty-based organization; the latter became the UK Government's implicit negotiation position when the realities of negotiations showed that the other analogies were unviable.
The postponement of de facto Brexit to 31 December 2020 reduced the size of the exit charge because of €18.5 billion being incurred between 30 March 2019 and 31 December 2020, during which time the UK will be treated financially 'as if' a Member State. However, this does not affect the money being paid over, as it constitutes a switch of headings from exit charge to net contribution over the 21-month transition period. A future risk to the UK is that the outturn exit charge will be higher if Contingent Liabilities, prominent in Table 2 (Barker), but missing from Table 3 (Office for Budget Responsibility), were to crystallize on a large scale.
Much negotiating conflict could have been avoided by an early UK offer to meet its net contribution for the last two years (2019 and 2020) of the 2014-20 MFF, including its share of RAL from 2021 onwards. This was conceded in Theresa May's Florence speech (May, 2017b), the EU27 then insisting on the transition period ending on 31 December 2020, the same day as the 2014-20 MFF. It is budgeting that really matters in the EU, where that is understood to include the MFF system and its RAL procedure (D'Alfonso and Sapala, 2015). The departure of a large net-contributing Member State has long-term implications for the fiscal sustainability of the EU budget and EU institutions more generally. By removing one of the most aggressive hawks on EU spending, Brexit may shift the balance of power away from Northern Europe, with conflicts over redistribution under MFF 2021-28 overlapping with those over threats to judicial and media independence in Eastern Europe. The notion of fiscal sustainability that shaped the EU negotiating position was a narrow one relating only to the 2014-20 MFF. In the background are the macroeconomic imbalances of the eurozone and tensions between Member States, particularly in relation to European Commission fiscal surveillance of Member States, using statistical accounting measures of deficit and debt.
Especially in times of austerity, access to public resources is determined by the budgeting mode, thereby explaining its centrality to political decision making: the EU emphasized the protection of commitments made in the 2014-20 MFF. The EU27 Member States supported the Commission, aware that shortfalls from Brexit would affect their statistical accounting data used for external fiscal surveillance. The calculation of the exit charge mostly derived from IPSAS-based financial reporting figures (Grossi and Soverchia, 2011), and notes to the accounts such as on contingent liabilities, on which the UK could make payments until 2064. Austerity has been required for reasons of fiscal sustainability following the 2008 global financial crisis and uneven growth within the eurozone, though measures taken to repair current deficits and reduce accumulated debt might damage long-term fiscal sustainability by depressing economic growth.
The exit charge has been an opportunity to explore the complementary roles of the four modes of government accounting in constructing an overall, fiscally transparent picture. An underlying problem is that all public spending numbers are large, and the lack of public understanding of their relative importance has a distortionary effect on public debates (Heald, 2012). The numerical significance of the net budgetary contribution and exit charge has been so exaggerated in UK politics that any number would have been politically toxic. Insiders know the fiscal irrelevance of the annual net contribution and exit charge compared to other likely effects of Brexit, yet some have been willing to accept far larger damage to UK public finances by risking a cliffedge exit with no agreement (that became known as 'No Deal').
The EU uses the 2012 Fiscal Compact 8 to tighten its fiscal control over Member States, including surveillance of contingent liabilities. Because the UK exercised its veto on treaty change, the Fiscal Compact was implemented through a 2012 intergovernmental treaty signed by all Member States except the UK and the Czech Republic. It represents a stricter version of the EU Stability and Growth Pact (European Commission, undated 2), accompanied by tighter enforcement by the European Commission. Yet, as Figure 2 demonstrates, the EU itself has developed off-balance sheet devices. The EIB is not consolidated in the accounts of the European Commission, and the European Fund for Strategic Investments (EFSI) is a joint venture between the European Commission and the EIB. Infrastructure projects can then be delivered in Member States through EFSI projects, including public- private partnerships that are designed to meet the criteria established by Eurostat (2016). This allows off-balance sheet treatment in statistical accounts, whatever the financial reporting treatment under IPSAS 32 (IPSASB, 2011). Turning to the second research question, our analysis shows that the contest between modes of government accounting is not just in the political environment but in technical practice. Combining insights from the four modes contributes strongly to fiscal transparency. Budgeting is central to authorization and legitimacy; financial reporting increasingly conforms to international standards; and harmonized statistical accounting provides the data for external fiscal surveillance (Heald, 2013). Although less widespread and inevitably model-based, fiscal sustainability reporting brings in long-term obligations that fail the recognition tests of financial reporting and statistical accounting.
This analysis emphasizes the importance of consolidation and the temptations that public decision makers face to put activities off-balance sheet. Bergmann et al. (2016) attribute the growing attention in OECD countries to consolidated government financial reporting to the increasing fragmentation of government, in part due to the influence of new public management. Consolidated information can provide an overview of the financial performance and position of government which the accounts of individual entities cannot do. As accruals-based government financial reporting takes hold, consolidation brings useful information about the 'whole picture' (Bergmann, 2014;Heald and Georgiou, 2011), bringing to the fore activities that would otherwise not be visible. Faced by such constraints, governments and international institutions seek off-balance sheet mechanisms to achieve policy objectives without the transactions being recorded as public expenditure or as public debt. Both financial reporting and statistical accounting are vulnerable, meaning that constant vigilance by standard setters is essential.
A warning to accounting standard setters and to public sector accounting researchers is that, in particular political circumstances, expert opinion can be trumped by lack of understanding and/or wilful misinterpretation of data. The technical accomplishment of government financial reporting changes has not been matched by success at communicating government financial performance. The questions of accessibility, intelligibility, and actual use should be high on the agenda of standard setters, governments, and public sector accounting researchers.

POSTSCRIPT
When this article was completed in November 2018, the UK was scheduled to leave the EU on 29 March 2019. Opposed by both 'Remainers' and hard-line 'Brexiters', including MPs in the Conservative Party, UK Prime Minister Theresa May was unable to pass the necessary legislation through the House of Commons to authorize the Withdrawal Agreement. The EU granted the UK an extension until 31 October 2019. She resigned as Leader of the Conservative Party on 7 June, staying on as Prime Minister until the result of the party leadership ballot was declared on 23 July. Having definitively committed to leaving the EU on 31 October, the newly elected Leader, Boris Johnson, became Prime Minister on 24 July. Unless the EU grant a 'better' Withdrawal Agreement by 31 October, which it has steadfastly refused to do, Johnson has committed himself to a 'No Deal' Brexit (i.e., exit without agreement) on that day, and has stated that he will withhold the exit charge.