Institutions and the route to reform of the European Union’s budget revenue, 1970–2017

Using process tracing, this paper charts the history of the changes in the EU’s revenue since 1970, including package deals and the unforeseen consequences of change, comparing the positions of the Council to those of the European Commission and European Parliament. Those revenue decisions allowed European integration to proceed though without a fully autonomous budget as Member States became more careful to calculate their net benefits or costs in relation to the budget. In December 2013, the European Union’s institutions established a High Level Group to recommend changes to the revenue base of the EU’s budget. This reported in January 2017, proposing to resolve the effect of sub-optimal revenue and budget decisions made by the European Union over many years, to reduce direct national contributions, to minimise the risk of unforeseen consequences, and to combine revenue flows with steering effects to discourage certain forms of economic behaviour in line with the wider policy agenda of the European Union.


Introduction
The budget of the European Union (EU) has always been highly contested, even before the United Kingdom (UK) had become a member in 1973. Indeed the European Economic Community (EEC) was financed under a system of national contributions (Article 200 EEC) that was phased out by the Own Resources Decision of April 1970, which prepared the way for the enlargement of the EEC to the UK.
In 2003, the Sapir Report called for a reduction in the 'relative weight of national contributions … in favour of revenue sources with a clear EU dimension ' (Sapir 2003: 166). The intention was to move away from the concept of juste retour or net balances that had become gradually entrenched since the UK had joined the EEC. It is net balances that lead wealthier contributor Member States to limit the budget, while protecting clienteles that benefit from redistribution such as the agricultural sector. This means that the EU cannot prioritise policy to meet collective needs or to respond rapidly to changing or unforeseen eventualities, in other words to provide a Europe-wide budget other than to the limited extent that applies to research and development (R&D), education, or Connecting Europe (the programme for communications and other network infrastructures) under Heading 1a, which accounts for 13 percent of expenditure, to freedom, security, justice, and citizenship, which amount to 3 per cent of expenditure, and to foreign policy priorities which amount to a further 6 per cent. The view of Sapir (2003) was that changes to the form of the EU's revenue away from contributions based predominantly on gross national income (GNI) would solve the problems of net balances and make the budget more responsive.
In December 2013, the EU's High Level Group on Own Resources (HLGOR), chaired by Mario Monti, was empowered to investigate and propose reforms to the EU's revenue. Reforms have been achieved in the past, starting in 1970. This paper evaluates those reforms that were always passed as package deals and affected expenditure and other policy commitments such as enlargement of EU membership. They also saw division about what sort of revenue should be raised and who would have control over it. The findings apply to the current period, with particular reference to the HLGOR, and to net balances given the British decision to leave the EU. The historic experience of the British rebate provides a lesson on package deals and the management of national issues in the budget.
The report of the HLGOR came at a time of challenge for the EU. The own resources agreements of 1970 and 1988 occurred when the bargaining power of the groups and states focused on the status quo came to be eclipsed by the power of a coalition for change, when the political and economic costs of the status quo exceeded those of the reform, and when small adaptive changes failed to ease pressure on the system (Lindner 2006: 169). The current state of health of the EU is one in which the political and economic costs of change may well be lower than perpetuating the net balance mentality.
This paper traces the path of the reforms to the EU's own resources from 1970 until 1988, and considers the current reform agenda. In doing so, it draws on process tracing (Checkel 2006(Checkel , 2008, and the historical institutionalist approach used by Lindner (2006) for analysing the EU's budgetary conflicts of the 1970s and 1980s. The paper applies these with regard only to the EU's revenue, with respect also to the positions of the European Commission and the European Parliament. Although their influence on the EU's revenue is limited, these latter institutions have the power to set agendas when revenue and expenditure reforms are linked. Finally, the paper explores the current reform agenda.

Challenges of own resources reform
The EU's common agricultural policy (CAP) and cohesion policy are much criticized for absorbing the bulk of the EU's expenditure and for the exercise in power for their protection since they defy the notion of public goods. However, conflict on the budget is due at least as much to the structure of its revenue side. This led in the first place to the creation of the British rebate in 1984, a system of smaller rebates for other countries, an explicit regard to national balances in the budget, and eventually to the departure of the United Kingdom (UK) from the EU largely on the basis of exaggerated claims about the UK's net balance.
The Multiannual Financial Framework (MFF) of 2014-20 fixed EU budget commitments at 1 per cent of GNI, although the absolute ceiling for own resources (or revenue) is set at 1.20 per cent of GNI. However, the value of the EU's financial flows exceeds 2 per cent of gross domestic product (GDP) when one considers the resources of the European Stability Mechanism (ESM), the European Development Fund (EDF), the European Fund for Strategic Investments (EFSI) and the trust funds for developing countries, all of which are either financed through intergovernmental instruments or by bond holders who are underwritten by national treasuries (Núñez Ferrer et al. 2016, chapter 3). The effect of the flows of these different instruments and of the de jure EU budget make the calculation of EU net balances spurious though they lead to the rebates. Even within the formal budget, some expenditure is traditionally redistributive and other spending has real multiplier effects, for example in the case of scientific research or the construction of networks that enhance energy security. Therefore, using the geographical location of beneficiaries to decide if expenditure is fair is a flawed approach. In the case of research expenditure, the benefit is not only financial nor limited only to the institution that wins the contract and performs the research.
Brexit, the euro area and the refugee crises provide demand for a budget to be credible for adapting to circumstance. Lindner (2006: 172) expects change to occur when its supporters increase their bargaining power, when the Council accepts reform and links its different components, which could include revenue and expenditure, when the status quo begins to be more costly than change, and when more modest amendments no longer satisfy. The circumstances of crisis offer an opportunity for reform as well as danger. In the past, change occurred when new actors had the power to overcome the status quo. This occurred with agreement on the British rebate in 1984 and in enlarging the budget with the GNI resource in 1988 (Lindner 2006: 174).
While the HLGOR published options for own resources in 2017, 40 years earlier, Donald MacDougall had led a similar enquiry that analysed public finances for the EEC (MacDougall et al. 1977). The report foresaw a tripling of the EEC budget from 0.7 per cent gross national product (GNP) to between 2 to 3 per cent within a 'pre-federal' phase. Such growth was justified if it could deliver value-added in the form of economies of scale, spill-overs or collective benefits, and fiscal neutralityor no overall increase to the fiscal burden (MacDougall et al. 1977: 14), characteristics of fiscal federalism (Blankart and Koester 2012). In 2017, the fiscal federalist emphasis on value-added pertained to economies of scale, threshold effects (to afford projects that would be unaffordable at purely national level such as the satellite or nuclear fusion programmes) and to cross-border externalities (Monti et al. 2017: 27). Fiscal federalism foresees that the federal level should provide macroeconomic stabilization (Alves and Afonso 2008: 20), defence, security and income redistribution unaffordable at the local level, while it should tax units that are mobile and able to avoid tax at local or state level, such that, 'central taxes can be more progressive [than local taxes], again without establishing fiscal incentives for relocation' (Oates 1999(Oates : 1128. Indeed the EU's internal market and inherent mobility allow opportunities for legal tax avoidance that fiscal federalism could reduce (Wyplosz 2015: 14).
In what follows, each of the major own resources reforms since 1970 will be analysed, looking at the package deals and the solution to previous crises. The last part of the paper will then evaluate possible reforms to the EU's revenue base.

The own resources decision of 1970
Own resources were agreed as a package deal in 1970, created to provide permanent financing for the EEC (Rittberger 2005). The package satisfied France -for the reasons of permanent financing of agriculture -and the other five Member States, which agreed to trade money for agriculture in exchange for French agreement to the EEC 's enlargement to the UK and (in the case of West Germany and the Netherlands) for the empowerment of the European Parliament (EP) over budgetary matters.
Until 1970, each Member State had to contribute resources from its national budget to the EEC to meet expenditure according to the following ratio set in 1957: Belgium 7.9; West Germany 28.0; France 28.0; Italy 28.0; Luxembourg 0.2; Netherlands 7.9 (Article 200.1 EEC). Concerning the European Social Fund only, the ratio was: Belgium 8.8; West Germany 32.0; France 32.0; Italy 20.0; Luxembourg 0.2; Netherlands 7.0 (Article 200.2 EEC). Unless there were unanimous agreement, updating the ratios according to economic or population growth or recession was impossible.
Article 201 EEC allowed for change in revenue, subject to everyone agreeing: 'The Commission shall study the conditions under which the financial contributions of Member States … may be replaced by other resources available to the Community itself, in particular by revenue accruing from the common customs tariff when finally introduced.' France wanted to stabilise financing of the CAP in order to avoid annual renegotiations. The financing of national contributions had been agreed to last until the end of 1968, with negotiations on a new mechanism starting in 1969. This coincided with the retirement of Charles de Gaulle from the Presidency of the French Republic. The European Commission proposed reforms to finances and procedures at The Hague European Council of December 1969 that were composed of a 'triptych' or package deal (Rittberger 2005: 199) made up of: achèvement of the common market in agriculture through a common financing scheme and the lock-in of the CAP; approfondissment of EEC relations through political cooperation in foreign policy and economic and monetary harmonisation; and élargissement to the UK.
The European Council in The Hague agreed on an own resources package that staged the introduction of a uniform external tariff and of agricultural levies between 1971 and 1973 as the EEC's sources of revenue (Conseil 1970a: article 3[2]), see Table 1. The budget would require full financing from the own resources of tariffs and levies from 1974 (article 5[1]) subject to unanimous approval by the Council after consulting the EP. Further resources could be transferred as a residual from Member States' taxes (article 5[3]). Transfers from national budgets in many Member States cause the collection of EU revenue to be entered as budgetary expenditure in national accounts, contributing to each country's deficit, a problem addressed in the report of the HLGOR (Monti et al. 2017: 25). The total of the resources agreed in 1970 was capped at 1 per cent of the EEC's GNP, reflecting a caution over budgetary expansion since before the UK became a Member State. Unwittingly, this allowed the UK to block future budget growth unless its own demands were met.
The EP tried to gain power over own resources through the imposition of amendments in what became the Decision of 1970, but failed. The EP wanted the package to go further both financially and procedurally. It proposed to give itself the power to assent or to reject. The EP also attempted to add the possibility of a residual resource based on a value added tax (VAT) call rate of 0.25 per cent from 1973 and 0.5 per cent from 1974 (article 4[3]). The Parliament proposed that from 1974, own resources could be amended or introduced not through national parliamentary ratification but by a unanimous decision of the Council together with a two-thirds majority in the EP (article 5b.1) while the tariff and VAT rates would be set by a joint decision of Council, EP and Commission. Whereas the Council had set total resources at no more than 1 per cent of GNP, the EP tried to change this so it could be increased via a two-thirds majority in the Council (article 5[4]). This would have prevented the UK from blocking an increase in resources in 1984 in order to obtain its rebate. Finally, the EP added a line to allow for resources to be generated from loans, which would have permitted the EEC to run a public debt (article 6). These ambitions of the EP are remarkable given that, until 1979, the EP was an assembly that represented national parliaments and reflected their pro-government majorities.
The Decision of April 1970 allowed for resources to be composed of tariffs, levies and VAT from 1971 with a phase-in period lasting until 1975, and with a VAT call rate of at most 1 per cent with rates set in the annual budget (Conseil 1970b: article 4). Each Member State was allowed to retain 10 per cent of the value the external tariffs collected. This amount increased in 2000 to 25 per cent but was reduced to 20 per cent in the Own Resources Decision of 2014.
Whereas the pre-1970 contribution system based on a ratio (Article 200 EEC) was that of an international, intergovernmental organization, tariffs and VAT as resources reflected the EEC's role as a customs union. Tariffs could affect consumer behaviour by making imported goods more expensive. VAT could also penalise consumers and this aggravated the UK after 1973, which had a large consumer economy relative to its economic size.

Treaty of Brussels 1975
The Treaty of Brussels in 1975 revisited some of the changes of 1970 in annual budgeting but not in own resources. The EP gained more power over the annual budget and the treaty established the Court of Auditors and extended the EP's audit powers. The EP unsuccessfully attempted to reform.
In its resolution of August 1975, the EP proposed joint powers with the Council in agreeing a procedure for annual VAT call rates to be fixed without regard for national parliaments (EP resolution C179/46 6.8.75, Section II, A9). Parliament urged its empowerment to co-decide own resources and upper limits with the Council (C179/46 6.8.75, III 21-2).
The Commission and the EP supported amendments to Article 203 EEC to allow the Commission to propose the rate of VAT in the draft annual budgets, for that rate to be set by the EP with a simple majority, and in the case of disagreement from the Council, for the EP to be empowered to overrule the Council and force through the VAT rate subject to a three-fifths majority. This would have allowed the EP to set the rate of VAT (an Own Resource) as easily as non-compulsory expenditure (Benedetto and Hoyland 2007). VAT was going to be the residual to make up for any shortfall from tariffs and levies whose total would not exceed the ceiling of own resources, at the time 1 per cent of GNP.
The Commission also proposed a new article 203b EEC, with the support of the EP, that would have permitted the raising of loans decided in the annual budget by the Council and the EP.
All of these own resources proposals were rejected by the Council in 1975. The proposals were radical and would have produced a very different set of budgetary financing arrangements if they had been agreed. In order to achieve better value for money through public goods, Fuest et al. (2015) recommend empowerment of the EP, whose interests are less particularist than those of the national governments on the Council, in budgetary expenditure. But not even they go so far as to suggest revenue based empowerment on this scale for the EP. The EP and the Commission were not proposing new own resources at this stage, but rather very significant powers for themselves in authorising the levying of resources.

Fontainebleau and the British rebate negotiations, 1983-1984
Besides the Brussels Treaty, 1975 was also the year of the referendum in which the British people decided to remain members of the EEC. Since that time, budgetary dissatisfaction by the British is chronicled for which temporary and suboptimal solutions were agreed when the willingness to do so existed (Lindner 2006: 128). The 1970 agreement had created a strong incentive for the EEC's post-1973 members to challenge the distributive and institutional status quo (Wallace 1983). By the early 1980s, the bargaining power of the UK had also increased (Lindner 2006: 186) given the need for its consent to raise the VAT call rate.
The end result from the Fontainebleau agreement was a package that included the permanent British ''correction'', an increase in the VAT resource to 1.4 per cent of the VAT base and agreement on enlargement to Spain and Portugal, which the UK might otherwise have blocked (see Table 2).
Initially the European institutions hoped to manage the situation of the UK's budgetary imbalance by insisting that spending was by policy rather than via transfers to Member States (Laffan and Lindner 2014: 228). The Commission's document of February 1983 refers to imbalances and proposed increased budgetary expenditure for the affected Member States by using a formula based on size of total agricultural expenditure (European Commission 1983a). Other concerns included enlargement to Spain and Portugal and the fact that the EEC was running out of money so an ambitious solution was needed.
The European Commission (1983a: 3) acknowledged the damage to cohesion in the EEC through the presence of imbalances. The UK rejected receiving extra expenditure as compensation and blocked any increase in the own resources' 1 per 'The prescribed ceiling of expenditure… could not be increased without the agreement of all member states… We could… make it plain that we were not willing to accept any increase in that ceiling… unless it was linked with an equally long-term reduction in the size of the British contribution' (Howe 1994: 306, emphasis in original, quoted by Lindner 2006.
The UK's blocking power was possible due to the ''Six'' having refused in 1970 to include the EP's proposal for the own resources' and VAT ceilings to be increased by a two-thirds' majority in the Council rather than unanimously.
The Fontainebleau agreement had the effect of cutting off Europe-wide own resources so that EEC revenue would be exclusively state-based. 'Fontainebleau accepted that, not European taxpayers, but national governments, provide the ownresources' (Lindner 2006: 177-179).
The Conclusions of the European Council (1984) at Fontainebleau guaranteed for 1984 a single lump sum payment to the UK of 1 billion ECU. From 1985, there would be a correction worth 66 per cent of the UK's over-contribution from the VAT resource. This would be paid one year in arrears through a VAT reduction for the UK paid by the other Member States through higher VAT contributions but with West Germany receiving a rebate on the rebate of two-thirds. The 66 per cent formula would only take effect once the VAT ceiling had been raised to from 1.0 to 1.4 per cent and would last only as long as the VAT call rate remained at least 1.4 per cent. There was a sunset clause in terms of conditions rather than time. If the VAT resource decreased, the rebate would expire, but the effect was to ensure rebate permanence with the UK in a position to protect the VAT ceiling from going below 1.4 per cent, unless the rebate were in future to be guaranteed by other conditions that the UK could also protect.
The Fontainebleau agreement was a package deal that allowed own resources to be increased, with the higher VAT call rate, the UK to be compensated, and enlargement to Spain and Portugal to proceed. However, the first draft of the 1985 Own Resources Decision proposed by the Commission neither mentioned the UK nor did it make the rebate permanent, with the Commission and the EP attempting unsuccessfully to resist rebate permanence.
In the original European Commission (1983b: article 3.3[1]) proposal there had been no reference to the UK, instead allowing for variable VAT rates for as long as agricultural spending under the European Agricultural Guidance and Guarantee Fund (EAGGF) and its successors accounted for more than 33 per cent of the EEC budget. This was an attempt to address British concerns over agricultural spending and, if adopted, would by now have seen the expiry of the British rebate as agricultural direct payments have indeed fallen to below one-third of EU spending. The rebate would have been 33 per cent (and not 66 per cent) of the net contribution and would have been amendable by a unanimous decision of the Council and a three-fifths majority in the EP without referral to national parliaments (European Commission 1983b: article 3.3[2-3]). This would have made future changes easier to achieve. The Council response (Council of the European Communities 1984: article 3.6) was to delete these proposals and put the Fontainebleau decision into effect. Besides a VAT call rate of 1.4 per cent, there would also have been to power to increase this to 1.8 per cent subject to a unanimous Council and a three-fifths majority in the EP, without referral to national parliaments. The VAT call rate would be set in annual budgets as a residual having taken into consideration expenditure and traditional own resources (Council of the European Communities 1984: article 3.2). The EP whose powers were only consultative attempted to resist this through an amendment to delete reference to the UK and to replace the rebate with extra spending using the following words: 'Whereas any Member State bearing an excessive budgetary burden in relation to its relative prosperity should, at the appropriate time, benefit from special Community measures in the fields of employment, energy and transport and any other suitable measures' (EP 1984).
The EP (1984: article 2c-d) also added that own resources could include VAT and anything else derived from a common EEC policy approved through Council unanimity and an absolute majority in the EP without reference to national parliaments. The concern of the EP was not to suggest new own resources but to gain the power to co-decide them with the Council.
Finally, the EP (1984: part B) added a second section to the legislation, subsequently rejected by the Council, which would have had the effect of imposing a three-year sunset clause on the UK rebate and requiring that, while the Commission would calculate the rebate each year, the ''correction'' itself would have been subject to the approval of the Council and the EP, giving the EP a veto on payment of the rebate.
In accordance with the Fontainebleau agreement, the Council deleted the Parliament's proposals and implemented the single lump sum payment to the UK of 1 billion ECU (Council of the European Communities 1984: article 8[3]), with the rest of the rebate coming into effect only upon ratification of the accession treaties with Spain and Portugal and the entry into force of the new VAT call rate at 1.4 per cent. The deal was packaged.

The own resources decision of 1988
Another package deal for own resources in 1988 used GNP, later GNI, to mobilise increased budget spending after the 1986 enlargement to Spain and Portugal and the Single European Act of 1987. The extra expenditure could put the single market programme into effect, providing a cushion through a doubling in size of the European Regional Development Fund. And yet, a budget expanded by the reintroduction of national contributions, on top of the rebates, reinforced a net balance approach by the Member States.
The 1988 Own Resources Decision was also necessary because the VAT call rate, which had increased in 1984 to 1.4 per cent, was insufficient and seen as iniquitous by less prosperous Member States where consumer spending was a larger part of the economy (Flaesch-Mougin 1986). Greater security in the budget with better permanent financing was deemed necessary, traditional resources were eroding due to lower tariffs and more self-sufficiency in agriculture, while the VAT call rate was not growing with the economy due to consumer reticence (European Commission 1987: 12). Own resources were needed that better reflected Member States' prosperity. Mobilising GNP also allowed for net balance calculations to be reinforced, providing a clear accountability to Member States like the UK.
The European Parliament (1986) report on future financing revisited the EP's previous rejection of the British rebate agreed at Fontainebleau, urged an increase in the VAT call rate, and the ability of the EEC to borrow for cohesion and growth. Other solutions for increased finance included increases in own resources through transferring national excise taxes to the EEC, even though excise rates varied hugely from country to country. A year later the EP (1987) urged the introduction of GNP percentage shares as an Own Resource to be collected not through direct national transfer but through the VAT call rate, while reducing the collection fee of Member States for tariffs from 10 to 5 per cent. GNP would provide for any shortfall, while its collection through VAT was hoped to minimise the net balance impact.
The Own Resources Decision of 1988 (Council of the European Communities 1988) reiterated the insufficiency of the 1.4 per cent VAT call rate, though retained it while capping it in cases where national consumer spending exceeded 55 per cent of GNP. It noted the Single European Act and the future need for stable revenue given the first Financial Perspective (or long-term expenditure programme) that took effect in July 1988 and was part of the package deal. The total ceiling for commitments in own resources was increased from 1.0 to 1.3 per cent of GNP.
The British rebate was retained and continued to be linked to the 1.4 per cent VAT call rate.

Reform challenges and opportunities
As the UK became more prosperous, the permanency of its rebate created a new iniquity in EU budget balances. The Commission and the EP in 1983-84 had attempted to avoid this by inserting sunset clauses or to allow the rebate only in the form of additional EEC expenditure. The British allowed the VAT call rate to be reduced and replaced by GNI percentage shares so long as the rebate were linked also to the GNI residual. Over time, the UK made incremental concessions to offset pressure for the rebate's abolition such as the exclusion from the rebate's calculation of non-agricultural expenditure in those countries that joined the EU since 2004.
As the UK's GDP increased, other prosperous net contributors demanded discounts though these all had sunset clauses attached to them that coincided with the expiry of the next MFF. The German rebate on the rebate has been extended from 66 to 75 per cent and also applies to the Netherlands, Austria, and Sweden. Several Member States receive a fixed correction (and unlike the UK, not subject to a complicated formula). For the period 2014-20, these amount annually to €695 million for the Netherlands, €185 million for Sweden, and €130 million for Denmark, while Austria received a total €60 million between 2014 and 2016. Payment of these is financed according to GNI share by all the other Member States including the UK. The VAT call rate has been successively reduced to 0.3 per cent, with a cap reduced from 55 per cent of GNP to 50 per cent of GNI, and a call rate reduced to just 0.15 per cent for Germany, the Netherlands and Sweden, amounting to a third rebate for these countries. All of these are incremental changes designed to contain the system by allowing net balance demands to spread from the UK to other Member States. Each micro-reform has led to further complication rather than simplification of the system (Ackrill and Kay 2006). Heinemann et al. (2008Heinemann et al. ( , 2010 have put much thought into a solution for the rebates based on a generalised correction mechanism (GCM). With the departure of the UK from the EU, a GCM to replace the rebates may be less relevant -and the Commission's reflection paper assumes that Brexit means no more rebates (Oettinger and Creţu 2017). Heinemann (2015) observes that abolishing net balances through changes to own resources, even if possible, would not remove political opposition to spending. Some actors and Member States are ideologically opposed to expenditure at the European level 1 (Heinemann 2015;Heinemann et al. 2008: 99). Osterloh et al. (2008: 444) note that, given the veto power of national governments, any proposal for an EU tax would be blocked unless compensatory mechanisms were introduced. Under these circumstances and given the political and economic benefits of ending the status quo and increasing the EU's budgetary responsiveness to the crises, changes may be more readily agreed. As I mention below, compensation could be made available on excessive gross contributions if the compensation is tied to expenditure on EU priorities.
Some of the ideas that emerged during the period that led to agreement on the UK rebate in 1983-1985 remain relevant. These include sunset clauses and linking corrections to changes in expenditure policy and wider policy outcomes, particularly with regard to the benefits of the internal market.
Along the lines proposed by the EP in 1983-1984, tailored rebates could also be tied to co-financing EU expenditure rather than being ''free money''. If a rebate is delivered in future not on net contributions but on excessive gross contributions generated from new own resources, for example from carbon taxes that exceed 1 per cent of the GNI of certain Member States, the rebate would be in the form of cofinancing of existing expenditure in conformity with EU policy. This would cohere with optimal tax theory (Mirrlees 1976: 354) that total welfare and revenue should be unchanged, that if a new policy would change behaviour, a second policy (in this case a rebate) should ensure total utility. A surcharge on carbon would be redistributed in the same Member State as investment in other policies. It is also desirable for production efficiency to be met through the tax system (Diamond and Mirrlees 1971: 277) by dealing with matters like carbon dioxide reduction with fiscal tools and in a fiscally neutral way.
8 Reforms: the High Level Group's report and the Commission's response, 2017 In December 2013, the MFF for 2014-2020 was agreed by the EU's institutions. As a condition for agreement, the European Parliament requested the appointment of a group to investigate new sources of finance for the EU's budget. In so doing, the Parliament maximised its agenda-setting power to try to escape regular disagreement on net balances of the Member States. That escape was contingent on reducing the weight of the GNI residual in the financing of the budget. The Parliament, the Council and the Commission agreed the appointment of a High Level Group to make recommendations by the end of 2016, while an interinstitutional conference attended by members of national parliaments would deliberate the reforms during the summer of 2016. 2 There are several candidates for new own resources. Their objectives should be to raise cash for the EU in a way that diminishes net balance arguments and to exercise steering effects to discourage certain types of economic behaviour and complement existing EU policy. Desirable steering effects for which an EU tax levy is the solution is a way to convince those who are otherwise sceptical about transfer of revenue to the EU level. MacDougall et al. (1977: 64) established criteria for future EEC financing that included yield, distributive capacities, economic function and administrative and political considerations, according to the type of integration that they could foresee: either 'pre-federal', federal with a small public sector, or federal with a large public sector (MacDougall et al. 1977: 19). Possible resources included, reflecting the world of 1977, an oil import levy, payroll deductions to finance EEC unemployment insurance, agricultural levies, and a fiscal complement to cohesion policy to tax regions with labour shortages and high incomes, the opposite criteria to the allocation of cohesion moneys.
An important responsiveness criterion for new own resources is that they take into account the mobility that the internal market and globalisation offer. Through the internal market, transnational corporations can avoid tax, putting smaller and medium sized competitors at a fiscal disadvantage. This type of iniquity could be resolved through an EU-wide corporation income tax (CIT) in order to mitigate against the race to the bottom (Schratzenstaller 2013: 310). A drawback is that the higher the mobility of a tax target, the higher the revenue volatility (Osterloh et al. 2008: 459). This makes retention of GNI transfers as a residual indispensable since the EU cannot incur debt.
An effective strategy is to look for new own resources where there can be political support for their steering effects besides CIT mentioned above. The roles of the FTT and carbon taxes are compelling in this regard, though they also face strong opposition. For this approach, the primary role of the FTT would be to discourage financial markets from making particularly risky transactions; that it may raise revenue is a secondary advantage. For Heinemann et al. (2008: 80), EU taxes are a matter of system design and choice of base. The EU could participate in harmonised bases across the EU in VAT, in a new EU CIT, or in the Emissions Trading Scheme (ETS), which could become an EU resource. In these cases, the EU could add its call rate via surcharges to differing national rates or it could attempt to charge the same rate across all Member States.
A carbon tax, aside from any EU consideration, is popular in several states in North-West Europe, where environmental concern is particularly high, a consistent pattern reported by Eurobarometer. For example, its spring 2015 edition (European Commission 2015: 19) revealed that 54 per cent of Swedes and 44 per cent of Danes believe that the EU budget should be spent on climate change and environmental protection before anything else, with 39 per cent of Germans, 35 per cent of Finns, and 33 per cent of Dutch and Austrians agreeing. A carbon tax based on energy usage or aviation would likewise be popular in North-West Europe and would be consistent with EU policy on climate change. It is likely to be unpopular in less prosperous Member States where the use of carbon in energy supply is proportionately larger. A solution is to cap Member States' gross contributions to a percentage of GNI. If the revenue from new own resources exceeds that gap, a correction would be supplied. The corrections would (as mentioned above) no longer be ''free money'', but would be tied to co-financing expenditure from the EU in order to protect the steering effect of the own resource, in this case reduction of carbon usage.
The HLGOR's report (Monti et al. 2017) evaluates a number of these possible new own resources. The priority of the HLGOR, however, was to explore methods for achieving a deal and addressing the issue of net balances so as to convince Member States to endorse a new type of budget whose revenue and expenditure would be Europe-wide and not national, and whose use could be more responsive to the sort of unforeseen events that the EU had faced over the previous decade. Table 3 summarises the nine recommendations in the report, which focus on the criteria and methods for reform rather than the reforms themselves. Recommendation 2 urges that the size of the budget should not be increased, and that own resources should support key EU policies in the internal market and with regard to environmental protection, climate change, energy union and transport (Monti et al. 2017: 11). Recommendation 3 (Monti et al. 2017: 12) affirms that Traditional Own Resources (tariffs) should remain and are a good model of a truly European own resource. It also urged conserving GNI percentage transfers as the stable residual for making up finances not provided from other sources. Recommendation 4 lists several possible new resources associated with better functioning of the internal market and fiscal policy: a new VAT resource, CIT, and Financial Transaction Tax (FTT) to reduce tax competition in the internal market. Energy and environmentrelated own resources would include a levy on carbon dioxide emissions, use of the Emissions Trading Scheme as an EU own resource, electricity and motor fuel taxes, and taxes on imported goods based on the carbon dioxide consumed in their production.
Recommendation 6 (Monti et al. 2017: 13) addresses notions of net balance, costs and benefits. It urges a way out from seeing the budget as zero-sum and in favour of European value added on investment in security, addressing climate change, and research. It includes benefits from such investment that are not directly financial, notably the gains from the internal market, the EU's role in global trade or climate negotiations. In this instance, the best forms of revenue are non-geographic and may, for example, include tax on a corporation's Europe-wide profits (the CIT) or on carbon use. Expenditure in one Member State may therefore not result in a ''loss'' by another Member State. The report also considers other bank levies besides the FTT and seigniorage, profits incurred by the European Central Bank. Annex VII (Monti et al. 2017: 87) assesses existing and potential own resources for efficiency, sufficiency and stability, transparency and simplicity, democratic accountability and budgetary discipline, European added value, subsidiarity, and reduced transactions costs. It finds that electricity taxes, Traditional Own Resources, motor fuel levies, an EU CIT, the GNI-based residual and an EU VAT score most highly for these criteria. Carbon dioxide levies, the Emissions Trading Scheme, bank levies and seignorage follow. Finally, the FTT and the current VAT resource score least well.
The Commission's reflection paper, issued 5 months after the HLGOR report, links own resources to EU policy priorities -so not only as revenue -and declares that Brexit solves the blockage of rebates (Oettinger and Creţu 2017: 9). It does not evaluate or recommend particular own resources. Consistent with Monti et al. (2017), Oettinger and Creţu (2017: 27) favour the principles of fiscal federalism with regard to value added and the pursuit of common values, mentioning investment in freedom, security and justice, a European Defence Fund, and public goods like Horizon2020, Connecting Europe, and the Satellite programme. They consider new resources to be positive without an increase in the size of the budget. Together with Monti et al. (2017), they consider potential options to include existing own resources, as well as a reformed VAT resource, CIT, FTT, electricity tax, motor fuel tax, seigniorage, and carbon pricing. Unlike Monti et al. (2017) they do not consider inclusion of the EU emissions trading scheme proceeds or a bank levy, but they do list the proceeds from the European Travel and Authorisation System (ETIAS), which will collect fees from non-EU/Schengen passport holders who enter the Schengen area. Oettinger and Creţu (2017: 29) favour the abolition of all rebates, particularly if real value added is achieved. This would appear to weaken package deal possibilities if a new Own Resource were to have a disproportional effect on a particular Member State. Like Monti et al. (2017), they favour policy links that achieve behavioural outcome, concerning energy or carbon pricing, while a CIT or FTT could reinforce the internal market and address tax evasion.  MacDougall et al. (1977) framed their considerations in terms of the EEC achieving pre-federal, or fully federal characteristics, the options presented by Oettinger and Creţu (2017: 30) are framed in terms of the EU either: 1. carrying on as before, minus the UK; 2. doing less but all together, 3. an EU in which some do more but others less; 4. a radical redesign, both reducing and increasing activities; and 5. doing much more all together. Each of these paths would require a different level of finance, which may draw on resources that also have steering effects.
The High Level Group's report and the Commission paper provide a route to a budget that is stable, agile and whose financing can influence policy. In the light of the EU's recent crises and Brexit, it remains to be seen if there will be a consensus in favour of such change.

Concluding remarks
This paper has analysed the evolution of the EU's own resources since 1970, using the principles of process tracing (Checkel 2006(Checkel , 2008. Agreements were reached as part of package deals on matters that included revenue increases with compensation for potential losers and the acceptance of the then EEC's enlargements. Those agreements drew on external tariffs, which had an intrinsic logic since the EEC of 1970 was a trading bloc whose main internal policy was provision of agricultural goods, and on consumer spending, again intrinsic to the market. Designing revenue with steering effects was not on the agenda. Permanent compensation for a named loser, the UK, had the unforeseen consequences of greater iniquity as the UK became more prosperous and elicited net balance demands from other Member States rather than defence of public goods. Unintended consequences of the 1970 decision also facilitated the UK's capacity to extract its rebate in 1984. In turn, net balance considerations generated by the UK rebate prevented the EU from adapting its revenue base to achieve steering effects in a changing social and economic environment. The use of sunset clauses and flexible compensation mechanisms may help to avoid unforeseen consequences in the future. Previous historic achievements on own resources were package deals, even the case of Fontainebleau in 1984, which allowed the VAT own resource to rise to 1.4% and for enlargement to Spain and Portugal. Coalitions for change can be very broad. While net beneficiaries could support the reduction of the GNI resource because it would make a beneficial budget more sustainable without having to fight net contributors all the time, net contributors can also gain if the package carries with it other policy achievements with the steering effects that they would support. At the same time, a reduction in the transfer of GNI percentage shares due to the uptake of new own resources leads to less pressure on national budgeting. The report of the HLGOR (Monti et al. 2017) and the study prepared for it (Núñez Ferrer et al. 2016) review a number of potential own resources for the future. These authors' work includes methods for addressing notions of net balance in order to arrive at a reform, as well as package deals as ambitious as those concluded in 1970, 1984 and 1988. New own resources can be non-geographic and can carry with them desirable policy-based steering effects to improve the functioning of the internal market, energy union, or climate policy. In the event of any national economy being unduly affected, for example less prosperous Member States dependent on carbon, temporary correction mechanisms can be put in place that do not endanger the steering effects of any new own resources.