Sir Phillip Green’s report to the PM on government procurement looks thin and is underwhelming in citing evidence. Yet Patrick Dunleavy argues that its key weaknesses essentially reflect the very poor state of the civil service’s budgetary and costs information. After two decades of ‘new public management’, Whitehall is light years away from the modern organization architectures of the world’s most efficient private corporations, which often stress ‘intelligent centre plus devolved delivery’ patterns. Green’s brief was efficiency, but his recommendations that government procurement become more integrated actually imply far-reaching governance changes across central government and beyond.
The Efficiency Review conducted in record time by Sir Phillip Green, the controversial boss of the Arcadia shopping empire, does not look like a conventional official report. Instead it resembles nothing so much as the kind of Powerpoint set that might be produced by a trainee member of the Cabinet Office’s Strategy Unit. Its meagre 33 large-print pages are structured in a randomish-looking sequence, with a couple of pages overview at the start and recommendations at the end, but with the middle being a magical mystery tour around different types of procurement. Most slides consist of just a few ‘Would you believe it!’ anecdotes. Of systematic data analysis, there is none.
Yet the report is also incredibly sharp on a key point – that its own deficiencies reflect major structural problems in how Whitehall now runs its business compared with the way that the most efficient giant corporations (like Walmart, Tesco, but perhaps not yet Arcadia) run their operations. These firms have a strong ‘intelligent centre’, bursting with information about their customers, and organizing strong nationwide procurement of thousands of goods (and now services also) based on the closest analysis of their datasets. Indeed, there is a case for arguing that thanks to its loyalty card Tesco, along with Google, probably knows more about UK citizens’ behaviours and views than government does. And the Tesco or Google data is real time, while much of the key government data is months or years old. Tesco can do centralized procurement for its 1,200 plus stores only because: (a) its expertise at the centre is so strong; and (b) it has also devolved delivery (dealing with customers, rostering in staff and last-minute co-ordination of deliveries of centrally-ordered products) to its store managers, leaving the centre free to do the big strategy picture.
By contrast, Green’s report makes clear that central government is set up to be a ‘dumb centre’ that knows little and has to go back to basics (i.e. manual returns of statistics) in order to understand the simplest cross-government function (like purchasing travel, mobile phones or even paper). He is repeatedly astonished at how little the senior civil service know about what Whitehall’s £191 billion of procurement spend is actually being devoted to. It quickly becomes clear that Green’s report had to make do with anecdotes because that was the best that the top civil service could come up with by way of cross-government information. Here are a range of the report’s comments on this aspect:
‘The inefficiency and waste of Government spending … is due mainly to very poor data and process’. (p.2)
‘Data is very poor and often inaccurate… There is no process for setting and challenging detailed departmental budgets’. (p.4)
‘Procurement data is shocking – it’s both inconsistent and hard to get at’ (p.7)
‘Central Government data is dependent upon each department submitting a manually produced return, or asking suppliers to provide cross-Government data… The quality of data provided by departments is inconsistent – there needs to be a central mandate to provide accurate and timely data’ (p.11)
‘The Government does not effectively monitor costs and efficiency at the centre. There is no effective mechanism to regularly challenge efficiency outside the department. Budgets are only reviewed by the centre at the highest level and rarely against any key performance indicators’ (p.31)
‘Reporting [of spending] should be standardized, to give a total Government view across categories’ (p.33)
Green’s recommendations are presented as being procurement common sense, but in fact they raise acute questions about the whole governance of Whitehall. He argues convincingly that central government should leverage its enormous buying power by acting as a coherent unit in its dealings with contractors, saving on repeating contract negotiations and achieving huge price discounts through aggregating demand.
Yet this is precisely the centralized approach that the civil service dismantled as woefully inefficient in the 1980s, 90s and early 2000s, and replaced with the still dominant but intellectually dead ‘new public management’ approach discussed elsewhere in two of my recent blogs. And the truth is that back in the 1980s, centralized procurement had many problems, including slow responsiveness, buying out of date computers at high prices, and adding months of delay to new initiatives and projects. Similarly in the 1990s President Clinton’s National Performance Review in American federal government was largely targeted on over-bearing central procurement controls.
But that was then and this is now. Largely unnoticed by Whitehall’s wholly non-tech (even anti-IT) mandarin class, the whole social organization of information in the private sector has been transformed by a dialectical mix of centralizing and decentralizing IT-driven pressures, brilliantly analysed recently by an LSE Centre for Economic Performance working paper. Green’s report makes clear yet again that the digital wave has largely by-passed Whitehall still – for if it had not the procurement data he sought would have been lying around in spades, not requiring special investigations to turn up nothing more than a few manually collated data points and a small stock of anecdotes.
Green’s report significantly does not once mention any particular unit in the Treasury or the Cabinet Office as being in charge of getting value for money in procurement. The Treasury has given up any pretence of doing detailed budgetary control, as the Green report makes clear. And the Cabinet Office is a by-word for ineffectiveness in the public management terrain.
Nor does Green mention at all the National Audit Office, whose role is centrally about obtaining ‘value for money’ in central government, and which has battled rather ineffectually and at a detailed level with departments about exactly the issues he covers – property management, procuring commodities and contracting for IT services. Having worked on six cross-government VFM studies for the NAO, it is clear to me that NAO is severely structurally limited in its ability to secure movement by the civil service towards an ‘information-rich centre’. It is hard to understate how strongly departments and ministers resist any such push that might impair their ‘ability to manage’ as they see fit.
Green clearly sees key changes towards more integrated procurement across central government as necessary, indeed inconceivable to not undertake. Yet at their root, almost all of the current data and information problems in Whitehall are governance problems – problems of who controls what and who sets what priorities. Rebuilding even a modest central capacity to overlook central government is not technically difficult given current ‘second wave’ digital-era governance’ organizational technologies and capabilities. But it will require sophisticated public management and the strongest possible push from David Cameron, George Osborne and the government centre to get any significant change against the departmental and agency baronies that new public management has allowed to strengthen and multiply.
Professor Dunleavy’s co-authored book, Digital Era Governance (revised edition, Oxford University Press, 2008), discusses government procurement of IT services in the UK compared with six other OECD countries.
Click here to see Professor Patrick Dunleavy’s BBC interview on this subject (WARNING: 42mb file).
Click here to respond to this article.