Christine Bell: A Modest Proposal: Power-sharing for the UK

Christine Bell

This blog argues that the UK is moving towards a form of consociational (political power-sharing) constitutional settlement. I consider whether embracing and developing its silent consociational elements could form the basis for a more coherent and stable new devolution settlement.

What is consociationalism?

Sometimes called ‘power-sharing’, Arendt Lipjhart has identified 4 classic elements to consociational government that can serve as a definition for our purposes:

  1. Grand coalitions which ensure that elites come together to rule so that minorities are not permanently excluded from political power
  2. Proportionality introduced at every level: through electoral systems using proportional representation and through the allocation of resources
  3. Segmental group autonomy, for example, through territorial division
  4. Minority or mutual veto, which gives each group ‘a guarantee that it will not be outvoted by the majority when its vital interests are at stake’ (Lijphart, Democracy in Plural Societies: A Comparative Exploration (1977), p. 118)

Consociationalism currently exists within the British constitution in two places. Most obviously, Northern Ireland’s devolution framework is fully consociational. Much less noticed the UK’s current devolution and Westminster arrangements exhibits a partial consociationalism with respect to the inclusion of different sub-state ‘national’ interests. A quick appraisal in terms of Lijphart’s criteria illustrates.

Proportionality. Lijphart’s second criterion of proportionality is met by the Westminster Parliament’s representation of sub-state regions on some sort of proportional relationship of the regions to the centre. Westminster has 650 MPs, proportion in the legislature – 533 English seats, 59 Scottish seats, 40 Welsh seats, and 19 Northern Irish seats. In terms of resources – another part of Lijphart’s proportionality – the Barnett formula attempts to reflect some sense of proportionality (and proposals to reform tend to be made on arguments for how proportionality should be calculated more accurately). The House of Lords has also traditionally been understood as brining an element of regional representation; and the three main political parties understood themselves as regional amalgams (apart from in Northern Ireland), although critically this has broken down in recent times, notably with relation to Scotland (adding impetus to my modest proposal).

Segmental group autonomy. Lijphart’s second criterion is also met, at least for three nations in their regions. Regional autonomy over cultural and social and increasingly economic affairs with a nod to national identity differentiations (whether politically or ethnically defined) is present through devolution of power to sub-state ‘national’ entities and legislatures in Scotland, Wales and Northern Ireland. Notably, however, one region and nationality, ‘England’ and ‘the English’, do not have devolved representation as such. However, ‘English votes’ for English laws in essence creates the UK parliament as a devolved legislature for England in its areas of vital interest.

Mutual veto on areas of vital interest. There are elements of Lijphart’s fourth criteria: areas of vital interest for sub-national units can be understood to be protected by effective quasi-veto through the devolution settlement, operating through legislative consent motions which prevent Westminster from unilaterally encroaching on devolved power of sub-state regions. New Scottish and Welsh Acts look set to take this further by providing a legislative basis for those legislatures as permanent (however binding that is).

Again, at present, this veto only operates with regard to Scotland, Wales and Northern Ireland. But as noted above, were the McKay Commission’s recommendations or similar to be adopted, then England too would have such a ‘veto’, albeit operating through an English sub-section of the Westminster Parliament, rather than a separate devolved unit.

Grand coalition? Crucially, however, one critical aspect of the consociational model appears to be missing: a clear commitment to grand coalition at the executive level, with an element of associated mutual veto. Although the UK has known ‘coalition government’, these were voluntary, partial and temporary rather than required, and did not have the aim of including the devolved regions at the centre, but rather of producing workable government at the centre in the absence of any one party winning a majority.

However, on closer examination even the executive is not entirely free from consociational tinge. Not only does each region have its own Secretary of State, but almost invariably every government is careful to ensure that the cabinet has Ministers drawn from Scottish or Welsh constituencies, although depending on electoral outcomes of course these may not be from the dominant parties in the devolved legislatures and regions.

D’Hondt is the mathematical formula whereby many consociational mechanisms (including Northern Ireland’s) translate proportions in legislatures into proportional entitlement to Ministerial seats at the executive. Interestingly, there is a handy website that does a quick ‘D’Hondt’ formula calculation. A quick run through of the last British parliamentary figures proves interesting. On the basis of a Parliament of 650, and the proportions of English, Scottish, Welsh and Northern Irish seats above, and the current cabinet size of 22, cabinet seats proportionally allocated by D’Hondt would amount to: England getting 19 Cabinet seats, Scotland 2, Wales 1, and Northern Ireland none. If one looks at the seats that the current cabinet hold, in fact this is close to how they divide: in terms of which regions the cabinet members are from: 19 are from England, 1 from Scotland, 1 from Wales and none from Northern Ireland (and one House of Lords member). (The last cabinet married exactly with D’Hondt, with exactly 19: 2: 1 and none, in terms of regional MPs who held office).

However, this form of representation itself does not mean that a classical consociational grand coalition exists: having the ‘right numbers’ of Welsh and Scottish representatives in the cabinet does not make the UK executive a consociational or power-sharing ‘grand coalition’. The current Welsh and Scottish ministers are chosen for their party affiliation and operate according to collective responsibility, rather than regional bartering. A truly consociational model would involve the grand coalition being compulsorily formed from across the lead parties in the devolved areas, and viewing themselves as on the cabinet to represent those areas.

Developing the UK’s current consociational model

Given this present partial consociationalism, could the UK model embrace and develop its consociational dimensions to deal with its current constitutional log-jams around regional devolution?

It would indeed be possible to move to a more fully consociational arrangement involving a power-sharing executive as a means of accommodating different regional identities and politics. Arguably, this could also resolve current and potential constitutional impasses and crises.

There are a few options. The government could move unilaterally to suggest that political parties holding the majority in the regions could nominate their Westminster leaders to the UK executive as a new model of ‘Secretary of States’ for the regions (it might be a problem for Sinn Féin depending on oath requirements, were Northern Ireland to be represented). It might be possible to expand the cabinet and have the parties suggest non-MPs nominated by majority regional parties, who would then be made members of the House of Lords and sit in the Cabinet on that basis (again this poses some problems for nationalist parties – both Sinn Féin and SNP).

One could be bolder: as the House of Lords representatives show, the cabinet is already somewhat of a ‘corporate’ group body where people are there not just by virtue of their election, but by virtue of ‘who and what they represent’ whether that is ethnic minorities, business interests, sub-regional representation, or other forms of expertise. Perhaps then other forms of ‘unelected’ (to Westminster) politicians could be included, for example, on the basis of their elections in devolved legislatures in the UK – something that arguably brings more entitlement to a place in government than appointment to the House of Lords? Secretaries of State, for example, could also be First Ministers of the devolved regions.

However, with but a small jump, there is another less ‘new’, way of achieving grand coalition in the Lijphartian sense: create a second executive. Now that EVEL looks set to be in place, the English in a sense have also a stake in the ‘mutual veto’ system of devolution, and the Westminster Parliament in-effect is now two Parliaments: an ‘English’ parliament and a ‘UK parliament’ (echoes of Dicey’s opposition to Home Rule here). As has been pointed out, while EVEL might re-balance the mutual veto imbalances of devolution for the legislature, it leaves untouched the imbalances of the double-function of the executive as an English and a British executive.

So why not just create another parallel grand coalition executive for UK-wide issues, leaving an English and Welsh executive for those regions?

In fact a quasi-such body already exists. Established by concordat, the little-known (publically at least) Joint Ministerial Committee comprises the First Ministers of the devolved regions (joint First and Deputy First in Northern Ireland’s case), abides by a form of collective responsibility (or at least confidentiality), and deals with areas:

where it is necessary to ensure uniform arrangements for relations between the UK Government and the three devolved administrations. In particular, broadly uniform arrangements need to apply to: handling of matters with an EU dimension; financial assistance to industry; and international relations touching on the responsibilities of the devolved administrations. (Devolution Memorandum of Understanding)

It could be a particularly ‘British’ constitutional solution to the problem of needing ‘two executives’ to parallel the two Parliaments, to view this Ministerial Committee as a second UK-wide cabinet for reserved powers. Westminster (with McKay’s tweak) would work pretty much as at present. However, the Cabinet would now be the English Cabinet with power over policy in England, while the Joint Ministerial Committee (or something very like it) would be the UK Cabinet (echoes of Home Rule’s ‘imperial parliament’ here). The UK Cabinet would only deal with modifications of the common core of reserved powers, and their fall-out for each other. Interestingly, parliamentary sovereignty as a concept (if not its reach) would be untouched. Parliament would be sovereign, but only in the areas where it had not already devolved power (as now) and the UK government would still be the UK government, just differently made-up.

Is this not incredibly immodest nonsense?

On first appearances this proposal may appear outlandish. However, current reality is that in practice on some of the most fundamental issues, such as the Constitution, foreign relations (including membership of the Council of Europe and EU), it is both legally and politically very difficult to change UK policy fundamentally without the consent of the devolved governments and even populations. So why not acknowledge this and the reality that these types of change need regional inter-negotiation if the UK’s current political settlement is not to be de-stabilised?

Other important matters, such as (at present anyway) UK welfare (apart from NI) and taxation of course can be changed unilaterally by the centre (with some exceptions in Scotland). However, if devolution continues apace then taxation looks likely to be much further devolved fairly fast (see Smith, Silk, and the Stormont House Agreement in Northern Ireland). Westminster control of taxation will then become more and more irrelevant to anyone but those residing in England. And devolution may also see changes to welfare too – it is devolved in NI (although Westminster makes NI toe the line through financial leverage), and may well be further devolved in Scotland and Wales (indeed there are forms of indirect devolution emerging).

Politically, a consociational model also sits well with the idea of increased English influence over England-specific law and policy. It enables Westminster to shape both – but for England only. Consociationalism can sit happily with English votes for English laws, and the two cabinet development may even be the logical conclusion of EVEL.

Unlike other proposals for constitutional revision, which require a big-bang re-invention moment, this proposal could be achieved by the mere strong political commitment of any UK government to agree the fundamental tenets of the Union with the devolved governments and to commit to only change them with the agreement of the membership of the Joint committee. This could all be as simple as a revision of the existing concordat, or a piece of legislation locking this commitment in legislatively (as per devolution itself, and perhaps national debt limits).

As I have demonstrated, we are not a million miles away from the model already. Given the UK’s stochastic approach to constitutional reform, and the almost accidental unravelling of all the things that might bind the Union (any agreed political and territorial community, welfare state, human rights act, common commitment to EU), embracing consociationalism might be more pragmatic and modest than any of the chosen and unchosen alternatives.

Christine Bell is Professor of Constitutional Law at the University of Edinburgh.

(Suggested citation: C. Bell, ‘A Modest Proposal: Power-sharing for the UK’, U.K. Const. L. Blog (13th Nov 2015) (available at