Impasse in a coalition – and how it should be resolved
New rules are needed where a Coalition Government finds it impossible to apply collective responsibility. The BGI believes that where one of the parties wishes to announce a policy which has not been agreed collectively the following rules should apply:
I. where the policy affects the interests of other departments their Secretaries of State should be consulted;
II. the Treasury must be consulted about any proposal with major cost implications; and
III. in making an announcement the Minister concerned should state whether the policy is supported by the whole Government or just by his own party.
Why do we need these rules? Collective responsibility has become an important constitutional principle in the UK because, without it, the citizen and others dealing with the Government, including foreign governments, find it difficult to tell which statements from the Government represent its policy and which do not. Also Parliament would find it difficult to hold the Government as a whole to account.
In order to give effect to collective responsibility important Government decisions are best made collectively, but the experience since 2010 suggests that the parties to a coalition may sometimes find it difficult to operate within the normal framework for collective decision-taking. The Coalition Agreement said that collective responsibility would apply “save where it is explicitly set aside” and the current Programme for Government lists five such subjects. However, there have been instances where one or other party has made announcements on other, non-specified matters purporting to represent Government policy only to be publicly criticised or even repudiated by its coalition partner. On occasion the parties have appeared to take markedly different lines on the same issue, for instance on the review of Parliamentary constituency boundaries, the future of welfare reform, press regulation, green energy and, repeatedly, anti-terrorism policy.
Despite the frictions within the Government evident from these public disagreements, the present coalition has proved remarkably durable. In 2010 many people saw it as a short-term expedient which might founder once the electoral interests of the parties diverged. In practice, despite sometimes strong disagreement and much public criticism of each other, it has been in the interests of both parties to avoid breaking up the Government. It may be that some degree of fractiousness is inevitable in coalitions.
It is highly desirable that where parties are considering coalition thought should be given at the outset not just to their joint policy programme but also to the processes by which they will resolve their differences in government and the consequences if they cannot agree. This is an area in which they could look to the civil service for help with deciding the ground rules. There is evidence that that the two parties thought about this issue in 2010: see, for instance, the emergence of the ‘Quad’ as a forum in which party-political differences could be addressed. It may be that the problem has been less the architecture than the willingness of both sides to adhere to it. The Lords Constitution Committee (in its 2013/14 report on the constitutional implications of coalition government) recommended that future coalition agreements should include a description of these processes and we agree with this.
Looking at the way in which coalition governments work in other countries, it is clear that practice varies a lot: thus, in Germany they have operated tightly with little if any public disagreement, whereas in Israel, for instance, coalitions tend to be very loose, even disorderly, with sharp public disagreements.
We think it best if in the UK the parties can agree to operate within the normal rules governing collective responsibility and decision-taking. These are most conducive to cohesive government and to well-judged decisions recognising that many issues cross departmental boundaries. They should not be set aside frequently or lightly. However, judging from the experience since 2010, it is probably unrealistic to expect rigid adherence to these rules throughout a full-term coalition government, especially once a General Election approaches. One party may develop proposals on which it insists which are simply unacceptable to the other. But also something turns on the nature of the agreement between the parties: if, for instance, it has been agreed that one coalition partner will have freedom of action in one policy area and the other in some other area, the normal clearance rules may not readily apply.
Even in these cases there needs to be some mechanism to avoid unforeseen collateral damage to other departments’ responsibilities. So there would need to prior warning of an intended policy and an opportunity for colleagues to draw attention to any difficulties for their own departmental interests, together with a mechanism for resolving these problems. And there would need to be enough time for proper consultation to avoid one party “bouncing” the other.
It is also essential that the requirement to consult the Treasury before any public commitments are made to policy proposals with major cost implications should be respected. This would be especially important where the costs stretch beyond the period covered by announced spending plans.
At times under the present Government it has not been clear whether announcements of new policy were supported by the whole Government or just by one of its constituent parties. So it is important that in cases where the Government has not reached a collective position on an issue this should be made explicit publicly.
The Lords Constitution Committee recommended that collective responsibility should only be set aside where the Cabinet as a whole agrees to this i.e. an agreement to differ. This seems to us to be very desirable but it may be unrealistic. At the very least, however, one would expect the two party leaders to have discussed any disagreement and there certainly should be no surprises. Our rules proposed above seem to us to be the minimum necessary to maintain effective government.