Why can't ministers and senior servants get along? A beginner's guide to bad blood, Universal Credit, blame, briefings and Civil Service Reform - Prof Perri 6, Chair in Public Management, Queen Mary

"War" has broken out, journalists say. Introducing a confrontation between Conservative MP and former police and criminal justice minister, Nick Herbert and former cabinet secretary, Lord (previously Sir Robin) Butler on BBC Radio 4’s “The Week in Westminster” on Saturday 30^th November, the Daily Telegraph’s Peter Oborne positively salivated with enthusiasm in telling us that hostilities between ministers and the senior civil service are at their fiercest for generations. Conservative bloggers such as Peter Hoskin are similarly excited.

So what is the dispute about and why does it matter? As with real wars, tensions between coalition ministers and the senior mandarinate have several causes. Four issues have come together.

The most immediate casus belli is Universal Credit, the government’s scheme for merging out-of-work benefits into a single payment. This has not been a teaching case study of successful policy implementation. The Major Projects Authority concludes that £140m has been “wasted” on information technology assets which will have to be written off (the department insists that some could yet be useful). So far there have only been local pilots covering a limited number of the benefits intended to be merged. But these have shown a number of problems. Although the government announced that the scheme will be implemented nationally by 2017 except for “some” recipients of Employment Support Allowance, this seems a very ambitious goal.

So the first issue is whether the civil service is to blame. In September, the Work and Pensions minister Mr Iain Duncan Smith seemed to be accusing the civil Service ITteam. By October, the allegation was being bruited in the press – presumed on the basis of briefings from within the coalition – that the Permanent Secretary, Mr Robert Devereux, had been asleep at the wheel. Precisely because he is a civil servant, of course, Mr Devereux can’t give his version in public unless a select committee asks him a question in a way that he can answer truthfully without compromising his duty of loyalty to his minister. And that’s a tall order.

      

We shan’t know for months or years who was most at fault for which aspects of the Universal Credit IT debâcle. Did Mr Devereux give the project as much attention as he reasonably could, given his department’s size and the number of its other projects? Were the junior technical staff not up to the job? Or the project managers? Or were Mr Duncan Smith’s instructions clear enough to enable them to produce a clear and robust project brief? Was the aspiration too big and complicated in the first place, as critics such as Colin Talbot of the University of Manchester argued? We don’t know, but I shall not be surprised if, by the time the papers are declassified, there is some blame for everyone involved because British government IT fiascos usually arise from all those problems at once, and more.

      

Second, the senior civil service is offended by criticism in the press based on briefings presumed to come from ministers. Much of Lord Butler’s broadside was a denunciation of just this. The current Cabinet Secretary, Sir Jeremy Heywood, is reported to have defended Mr Devereux to the prime minister and, apparently, to have pinned the blame for the Universal Credit fiasco on Mr Duncan Smith instead.

      

This looks like the senior civil service briefing off-the-record in the manner that Lord Butler complains that ministers have been doing. But off-the-record briefing is a symptom, not a cause of the breakdown of relations.

      

By the time Mr Herbert squared off against Lord Butler, the issue seemed to be whether this case was symptomatic of a wider civil service inability to manage big projects. That brings us to the third issue. On “The Week in Westminster” Mr Herbert claimed that because no permanent secretary had resigned over any of the long series of big project failures from the last two decades, the senior civil service is not sufficiently accountable, and that by contrast any private sector chief executive who presided over a big fiasco would have walked the plank.

      

In fact, private company CEOs certainly don’t always resign when project failures come to light. It helps of course that when many corporate IT projects do fail, as they often do, CEOs rarely face a hostile media or any questioning at annual general meetings even faintly reminiscent of a grilling by the Public Accounts Committee.

     

In any large organisation, large projects often fail, have to be abandoned, are delivered late or over budget, produce disappointing results or have their scope drastically reduced to get implemented at all. The more fundamental question is whether we think that the most powerful incentive we can give top civil servants to get them to run big projects better is a blunt expectation that they must resign when things go wrong. One outcome of using that strategy with the senior civil service might be that we’d get resignations so often that the capability of top management would be undermined. Another is that civil servants would do everything they could to reduce the number of big projects or ambitious activities they take on. Ministers who complain that the civil service is risk averse ought to be careful what they wish for. A duty to resign for any project failure could make that problem worse. Alternatively, if ministers insist on big, high risk projects and then insist that top civil servants take the blame they go wrong, they can hardly be surprised if few really talented people want to become permanent secretaries. Putting the word “scapegoat” in a job description is not usually recommended by recruitment consultants. Even then, it isn’t very likely that ministers would escape blame by such a rule anyway. It would strain civil servants’ loyalty under our “public service bargain” too far to hope they would not respond by briefing the press that the fault lay with ministers.

      

Probably the fourth issue behind the current row worries the top civil servants most. Cabinet Office minister Francis Maude set out proposals in June for ministers have French style “cabinets” of politically appointed staff with powers to give orders to civil servants. Whether that would apply to departments such as her Majesty’s Revenue and Customs which don’t now even have their own minister remains to be seen. For a system which has, since the 1870s, relied on a politically impartial and permanent senior tier of officials, this would be a major constitutional change. For senior civil servants, reporting to a politically appointees would mean, they fear, the final demise of their already eroded role of being the main source of policy advice to ministers, given directly and without intermediaries. At the same time, Mr Maude’s plan would make permanent secretaries, not ministers, individually accountable to parliament for the implementation of major projects. Taken together, the two measures would seal the change in the senior civil servant’s role from that of policymaker to that of managerial agent.

      

The senior civil service has sought to defend its roles in policymaking and policy advice against efforts by governments since the 1980s to focus the service on management and “delivery”. One of the Thatcher government’s aims for the “Next Steps” programme was to develop a cadre of senior civil servants who would gain the highest status through management rather than policymaking, but the most ambitious continued to pursue policy work. The efforts of Sir Michael Barber and his team working for Blair’s administration to concentrate top civil service minds on “delivery” may have been unduly narrow and mechanistic in approach, but civil service focus on policy work was probably an equally important reason for their limited impact on civil servants’ priorities. When the coalition formalised the already widespread use of policy advice from think tanks and management consulting firms into a programme and a set of contracts, among least some senior civil servants seem to have responded with chagrin or disdain.

      

 In the United States, France and Sweden, a tier of ministers’ political appointees has powers to instruct civil servants. Since none of those countries has collapsed as a result, no one can claim that ending the century and half “British tradition” must necessarily be a catastrophe. Indeed, it’s a moot point whether we can something a “British tradition” at all, which itself took twenty years of effort after the 1854 Northcote-Trevelyan Report to get established. The commitment to a permanent civil service not reporting to political appointees was itself considerably overridden when Lloyd George brought in such personally appointed “men of push and go” as Sir Eric Geddes after 1916. Since the outbreak of war in 1939, increasing numbers of “irregulars” have been brought in to British government as temporary civil servants – more or less political appointees but given civil service status so that they issue orders to civil servants. Maude’s plan would turn them from irregulars into a different kind of “regular”, and would give them seniority over the routine kind of “regulars”.

      

Whether a ministerial “cabinet” with authority over the permanent officials would give ministers what they say they want is another matter altogether. Off-the-record briefings to the press in recent years suggest that ministers suspect the top civil servants of not being committed to implementing their policies. Even if that’s true – and of course it’s disputed by the civil servants – it’s far from certain that making them report to political appointees will make them more motivated. For all the merits of the French and US federal executive systems, there’s no evidence that they exhibit greater (or, admittedly, lower) levels of enthusiasm or commitment among civil servants for carrying out instructions. On the other hand, in Britain, losing opportunities to offer advice to ministers about feasibility of implementation or about scheme design – because ministers would take advice mainly from their own appointees – might make top civil servants less motivated rather than more.

      

Again, whatever the merits of ministerial cabinet system, there is little evidence that, in and of itself, it results in better design and execution of major projects. That France has had fewer disasters with government IT projects than the UK seems to due to quite other factors. And the fate of the Obamacare web portal reminds us that the US is capable of fiascos with its most ambitious schemes for online integration too.

      

A system of politically appointed senior executives in government is only as good at policy development and the oversight of civil service execution of programmes as the political appointees are. US government has sufficient resources to attract seasoned people to be political executives who have major public management experience as state governors or big city managers or who have worked in previous federal administrations. And the political parties can draw upon a large pool of experts in each ministry’s policy area who are aligned with them. The French system today – to the chagrin of many French critics – draws on a much narrower pool of talent. Tory, Labour and especially Liberal Democrat ministers in Britain would face difficulties in recruiting cabinet staff who are party loyalists, who understand the civil service but are not of it, who know what the implementing agencies and authorities can do, who understand how cross-departmental working in Whitehall has to be achieved, who can appreciate what costings are realistic, but who also have the political savvy to provide ministers with adroit political advice. Rather few people with all those characteristics are to be found in London’s think tanks or in party headquarters. Some might be found in local government, but not enough to meet the requirement.

      

The present senior civil servants say that they fear that politically appointed cabinets will be echo chambers, not places where even the political appointees will “speak truth to power”, let alone the permanent civil servants who are not invited. The evidence from other countries is mixed: toleration for challenge, use of devil’s advocates, interest in evidence and practice of open-minded deliberation vary widely between governments and between ministries in France and the US. Since Lord Butler complained on “The Week in Westminster” that even now under the coalition, civil servants are punished for speaking truth to power, even in private, one wonders whether political appointees would really be much more cowed than civil servants may be today.

      

But perhaps a more interesting question is how or how far Mr Maude’s system might address the demand from Mr Herbert for “accountability”, when that is understood as requiring resignation when things go wrong. If senior civil servants are to be asked to go when projects fail, under what circumstances will the tier of political appointees themselves resign, after being found to have advised a minister on a misconceived scheme?

      

As with most things in government, the first decision should be to choose which problem ministers most want to solve.

      

If we want to reduce the number of big project fiascos, it would be better to begin with a serious discussion about the issues of just how big and how complex a plan for a project has to be, before we can reasonably expect that it will probably go badly and ministers ought to be asked for something more practicable. Then we should look again at the organisational capability of project management in the civil service rather than endlessly fiddling with the sequence of approval stages for plans.

      

If we want to do something about civil service commitment, then we should come up with a set of explanations of just where, why and how far a deficit of commitment is important in explaining weak performance, and where, why and how far problems are due instead to poor management systems. Of those weak systems, one should be addressed as a priority. Few civil servants working on major projects work on them from beginning to end, whereas in local government it is much more common for the same managers to carry responsibility throughout a project’s life. The civil service moves them on sometimes after only a few months. The same has too often been true of ministers themselves. Until both ministerial and civil service career moves are more closely tied to project and programme achievement, project management will not be greatly improved just with improved techniques or more oversight at the various approval stages.

     

But if on the other hand, we really want to find a way to reduce ministers’ exposure to blame whenever things do wrong and instead transfer that exposure to civil servants, then perhaps we ought first to ask how we expect to adjust the “public service bargain” at the same time to offer some positive incentive for talented people to want to do senior jobs in the public service. More sticks and sermons without carrots may make for good headlines but usually make for poorly motivated organisations.

     

But the present “war” in the media so far suggests that may be all we shall get.

I am very grateful to Colin Talbot and Janice Morphet for their comments on an earlier draft of this post. Of course, neither bears any responsibility for my mistakes, nor should they be presumed to share my views.